JP7530630B2 - Testing systems and programs - Google Patents

Description

The present invention relates to an examination system and program configured by a computer that administers tests to examinees in remote locations via a network, and can be used to administer a variety of tests, such as qualification tests to measure abilities and knowledge such as language ability, conversation ability, communication ability, and skills, certification tests, qualification tests, entrance exams, employment exams, graduation exams, completion exams, final exams, promotion exams, mock exams, and daily learning effectiveness check tests.

In recent years, online learning instruction has become popular, whereby instructors and students converse over the Internet to learn languages and other subjects. In addition to language instruction for students, online courses aimed at acquiring various skills and knowledge necessary for finding a job or living in society are also becoming more common.

On the other hand, exams are necessary to measure the level of achievement in language learning, the degree to which knowledge has been retained, the degree to which skills have been acquired, etc. Exam results are directly linked to success or failure of, for example, the acquisition of qualifications, employment, enrollment, graduation, completion, progression, and promotion, so they must be administered rigorously and impartially. However, it is also true that in the past, cheating practices such as proxy exams, in which a third party other than the examinee takes the exam in place of the examinee, and exam-takers answering questions based on information provided by a third party during the exam, have been widespread.

In the present invention, the test is performed using a head mounted display (hereinafter, sometimes abbreviated as HMD). In terms of information processing using a head mounted display, a technology related to the present invention is known that allows a wearer of a non-transparent head mounted display to easily grasp the surrounding situation even when viewing video content on the non-transparent head mounted display (see Patent Document 1). This non-transparent head mounted display includes an imaging unit that captures an external image, an external image acquisition unit that acquires the external image captured by the imaging unit, a video content acquisition unit that acquires video content, a display unit that displays at least one of the external image acquired by the external image acquisition unit and the video content acquired by the video content acquisition unit, an approach detection unit that detects the approach of an object, and a display control unit that displays a superimposed image on the display unit in which the external image and the video content are superimposed so that both the external image captured by the imaging unit and the video content can be viewed when the approach of an object is detected while the video content is displayed on the display unit. The video content may be any content that includes images and audio, and is, for example, a movie, a television program, a game, etc.

A system is also known for transitioning between a transparent mode and a non-transparent mode in a non-transparent head-mounted display (see Patent Document 2). In this system, at least a portion of an image of the real world is displayed on a portion of the screen that renders an interactive scene of a game. In this case, the transition from the non-transparent mode to the transparent mode is made in response to a change in the line of sight of the user wearing the head-mounted display.

JP 2015-191124 A (Abstract) Special table 2016-532178 publication (summary)

As mentioned above, online learning instruction and online courses using the Internet are becoming more and more common, and as an extension of this, it would be convenient if exams could also be conducted online using the Internet, and there is a demand for this. In this case, exams test an individual's abilities and knowledge, and the results of exams are directly linked to the success or failure of, for example, the acquisition of qualifications, employment, enrollment, graduation, completion, advancement, and promotion, so exams must be conducted rigorously and without cheating.

However, in traditional exams where a large number of test takers gather in an exam hall or classroom, exam monitors and teachers patrol the exam hall or classroom to prevent cheating, but with online exams, such patrols are not possible.

Therefore, it is desirable to develop a system that can administer online tests to test takers in remote locations using the Internet while preventing cheating.

The object of the present invention is to provide a test system and program that can conduct online tests while preventing cheating.

The present invention provides an examination system configured with a computer that administers an examination to examinees in remote locations via a network, comprising:
A test management server is provided which is connected to an examinee terminal operated by an examinee via a network,
This test management server is
a question data transmission means for executing a process of transmitting the test question data to an examinee terminal via a network in order to display the test question data on a display unit of a non-transmissive head mounted display worn by the examinee;
an answer data receiving means for receiving the answer data of the examinee transmitted from the examinee terminal via the network;
a wearing/removing state data receiving means for receiving wearing/removing state data of the non-transmissive head mounted display, the wearing/removing state data being detected by a wearing/removing detection device provided in the non-transmissive head mounted display and being transmitted from the examinee terminal via a network;
and a fraud determination means for executing a process to determine whether the attachment/detachment state is appropriate for the question type of the question, using the attachment/detachment state data for the question time period and the attachment/detachment state data for the answer time period received by the attachment/detachment state data receiving means.

Here, the "process of transmitting test question data via a network to an examinee's terminal in order to display the test question data on the display unit of a non-transparent head mounted display worn by the examinee" in the "question data transmission means" includes a case in which the test question data is transmitted via a network to the main body of the examinee's terminal (a device independent of the non-transparent HMD) and then transmitted from that main body to the non-transparent HMD, and a case in which the test question data is transmitted directly to the non-transparent HMD via a network.

In addition, the "means for determining whether or not the attachment/detachment state is appropriate for the type of question asked" in the "means for determining whether or not the attachment/detachment state is appropriate" does not necessarily mean that a determination must be made as to whether or not the attachment/detachment state is appropriate for all questions, but rather that a determination must be made as to whether or not the attachment/detachment state is appropriate for questions that require the wearing of a non-transparent head-mounted display. Therefore, for questions that do not require the wearing of a non-transparent head-mounted display, a determination must be made as to whether or not the attachment/detachment state is appropriate. In addition, if all questions in a certain test require the wearing of a non-transparent head-mounted display, a determination must be made as to whether or not the attachment/detachment state is appropriate for all questions in that test.

In the test system of the present invention, questions are displayed on a non-transparent head-mounted display, so no one other than the examinee can see the questions. In addition, because the examinee is wearing a non-transparent head-mounted display, it is an environment in which it is difficult for the examinee to cheat, for example by secretly looking at a dictionary.

Furthermore, the wearing/removal state data detected by the wearing/removal detection device provided on the non-transparent head mounted display is used to determine whether the wearing/removal state is appropriate for the type of question posed, so that it is possible to capture, for example, a situation where a test-taker is answering a question that requires wearing a non-transparent HMD without wearing the non-transparent HMD, or a situation where a test-taker is not wearing the non-transparent HMD when they should be wearing the non-transparent HMD to view the question. This makes it possible to conduct the test while preventing cheating, thereby achieving the above-mentioned objective.

<Configuration for executing warning display processing on supervisor terminal>

In addition, in the above-mentioned test system,
a monitor terminal connected to the test management server and displaying information for a monitor to monitor the test;
The test management server
a monitoring means for receiving image data including the face of an examinee captured by a person imaging device provided in the examinee terminal and transmitted via a network, and for executing a process of displaying the received image data as a moving image on an observer terminal;
The fraud determination means is
During the processing of a question that requires the wearing of a non-transparent head mounted display, if it is determined that the wearing/removal state data during the question time period indicates that the display is being worn and the wearing/removal state data during the answer time period indicates that the display is not being worn, a process is executed to transmit information to the monitoring means that there is a possibility of cheating;
The monitoring means are:
It is desirable that the monitor terminal is configured to receive a signal from the fraud determination means and then execute a warning display process around or in a peripheral area of the image data of the examinee that is displayed on the monitor terminal.

In this way, when the monitor terminal is configured to execute the warning display process, the monitor who is monitoring multiple examinees can be notified of examinees who may be committing fraud, which reduces the monitor's monitoring burden and improves monitoring efficiency.

<Configuration for extracting facial features from each of multiple facial image data and determining whether all of the multiple facial features match or nearly match>

In addition, in the above-mentioned test system,
The test management server
a face image acquisition means for acquiring face image data from person image data including the face of the examinee captured by a person image capture device provided in the examinee terminal and transmitted via a network at the start and/or end of the test, and during the time period during the test when the wearing/removal status data indicates non-wearing;
a facial feature extraction unit that executes a process of extracting facial features from each of the plurality of facial image data acquired by the facial image acquisition unit;
a certificate generating means for executing a process of generating an electronic or paper certificate to be issued to an examinee;
The fraud determination means is
It is configured to determine whether or not all of the facial feature amounts extracted by the facial feature amount extraction means match or substantially match, and if all of the facial feature amounts do not match or substantially match, to execute a process of determining that fraud has been committed,
The method for creating the certificate is as follows:
If the fraud determination means determines that all of the multiple facial features match or nearly match, it is desirable to configure the means to execute a process of adopting the facial image data among the multiple facial image data which most clearly shows the examinee's face as the facial photograph to be printed on the certificate.

Here, "at the start of the test" in "at the start of the test and/or at the end of the test, and during the time period during the test when the wearing/detachment status data indicates non-wearing" in "face image acquisition means" means the stage before the process of sending and receiving question data and its answer data starts, and includes before the start of the test, and has a certain time span. However, including before the start of the test does not mean, for example, including the day before or several hours before, but means the preparation stage before entering the routine of seeing, hearing and answering questions. Similarly, "at the end of the test" means the stage after the process of sending and receiving question data and its answer data ends, and includes after the end of the test, and has a certain time span. However, including after the end of the test does not mean, for example, including the next day or several hours after, but means the time period that can be said to be immediately after. The same applies to the following inventions.

In this way, by extracting facial features from each of the multiple facial image data and determining whether all of the multiple facial features match or nearly match, it is possible to determine whether people have been replaced during the test and detect fraud, as well as obtain an appropriate facial photograph to include on the certificate.

<Configuration for issuing instructions to remove non-transparent head-mounted display>

Furthermore, in a configuration in which a facial feature amount is extracted from each of the plurality of pieces of facial image data described above, and whether or not all of the plurality of facial feature amounts match or substantially match is determined,
The test management server
a facial image acquisition necessity determination means for executing a process of transmitting an instruction to remove the non-transmissive head mounted display to the examinee terminal via a network at any timing in order to obtain facial image data for the time period during the test, when questions requiring the wearing of a non-transmissive head mounted display occur consecutively for a predetermined number of questions or for a predetermined period of time, and when it is determined that the wearing/removal state data during the consecutive questions indicates that the non-transmissive head mounted display is being worn;
The face image acquisition means is
It is desirable that the configuration be such that when the facial image acquisition necessity determination means sends an instruction to the examinee terminal to remove the non-transparent head mounted display, and when it is determined that the attachment/detachment status data after the removal instruction indicates that the display is not being worn, a process is executed to acquire facial image data from the person image data sent from the examinee terminal via the network.

When configured to issue an instruction to remove the non-transparent head mounted display in this manner, it is possible to obtain the necessary facial image data and detect cheating even in situations where it is difficult to obtain facial image data during the test, when there are successive questions that require the wearing of a non-transparent head mounted display. In addition, since the instruction to remove the non-transparent HMD is sent at any timing by the facial image acquisition necessity determination means when certain conditions are met, it cannot be predicted on the examinee's terminal side, and therefore, for example, it will be impossible for the assistant to detect the issuance of the removal instruction and return to the examinee, creating an environment in which cheating is difficult.

<A configuration in which the wearer's gaze image data captured by the wearer's gaze imaging device is displayed on a non-transparent head-mounted display and also on the monitor's terminal>

In addition, in the case where the monitor terminal is configured to execute the warning display process,
The non-transmissive head mounted display is provided with a wearer's line of sight imaging device capable of imaging a space outside the non-transmissive head mounted display with the line of sight of a wearer of the non-transmissive head mounted display,
The problem data transmission means is
During the processing of a question that requires wearing a non-transmissive head mounted display and that requires an answer involving a movement, at the timing of transition from a question period to an answer period, a process is also executed to transmit, via the network to the examinee's terminal, an instruction to switch to a display mode for answers involving a movement, in which image data of the wearer's gaze captured by the wearer's gaze capturing device is displayed as a moving image on at least a part of the display section of the non-transmissive head mounted display;
The answer data receiving means includes:
The device is configured to execute a process of receiving wearer's gaze image data transmitted from an examinee's terminal via a network as answer data,
The monitoring means are:
It is desirable to configure the device to also execute a process of displaying the wearer's gaze image data transmitted from the examinee terminal via the network as a video on the monitor terminal by switching between or together with the display of the person image data.

Here, "questions that require an answer involving an action" include, for example, test questions to be answered in writing, practical test questions on sign language, practical test questions on how to use tools, equipment, and the like, test questions on how to tie ropes on a ship, etc.

In this way, when the wearer's gaze image data captured by the wearer's gaze imaging device is configured to be displayed on the non-transparent head mounted display and also on the monitor terminal, when a question is asked that requires an answer involving movement, the examinee can answer the question while checking the movements of their hands, fingers, etc. with their own eyes, even while still wearing the non-transparent head mounted display. In addition, by displaying the wearer's gaze image data on the non-transparent head mounted display, the examinee can see the outside of the non-transparent head mounted display, which creates an environment in which cheating is likely to occur; however, since the wearer's gaze image data is also displayed on the monitor terminal, cheating can be prevented in advance.

<Configuration that performs character reading processing to evaluate written answers>

Furthermore, in a case where the wearer's gaze image data captured by the wearer's gaze imaging device is displayed on a non-transmissive head mounted display and also on a monitor terminal,
The test management server
It is preferable that the device further comprises an evaluation means for executing character reading processing using the wearer's gaze image data received as answer data by the answer data receiving means for questions to be answered in writing.

In this way, if the written answers are evaluated by performing a character reading process, the evaluation of the examinee's answers (determining whether they are correct or not, marking, etc.) can be automated.

<Configuration that performs pattern recognition processing to evaluate answers given by actions in real space using actual test equipment>

In the case where the wearer's gaze image data captured by the wearer's gaze imaging device is displayed on the non-transmissive head mounted display and also on the monitor terminal,
The test management server
The test device includes an evaluation unit that executes a pattern recognition process using the wearer's gaze image data received as answer data by the answer data receiving unit for questions that are answered by actions in a real space using real test tools,
This evaluation tool is
It is desirable to configure the system so that pattern recognition processing is performed using a model for pattern recognition obtained by a learning process using wearer gaze image data that has been determined to be correct and wearer gaze image data that has been determined to be incorrect, as well as the corresponding correct and incorrect labels.

In this way, if a pattern recognition process is performed to evaluate answers based on movements in real space using actual test equipment, it is possible to automate the evaluation of examinees' answers (such as determining whether they are correct or not and marking them).

<A configuration in which the wearer's gaze image data is displayed on a non-transparent head-mounted display and also on the monitor's terminal so that the examinee can take written notes>

In addition, in the case where the monitor terminal is configured to execute the warning display process,
The non-transmissive head mounted display is provided with a wearer's line of sight imaging device capable of imaging a space outside the non-transmissive head mounted display with the line of sight of a wearer of the non-transmissive head mounted display,
The problem data transmission means is
During the processing of questions that require the wearing of a non-transparent head mounted display and for which written notes are permitted during the question setting and answering periods, test question data is displayed on the display unit of the non-transparent head mounted display, and a process is also executed to transmit to the examinee's terminal via the network an instruction to switch to a memo display mode that displays wearer's gaze image data captured by the wearer's gaze imaging device so that the examinee can take notes in writing,
The monitoring means are:
It is desirable to configure the device to also execute a process of displaying the wearer's gaze image data transmitted from the examinee terminal via the network as a video on the monitor terminal by switching between or together with the display of the person image data.

In this way, if the wearer's gaze image data is displayed on the non-transparent head mounted display and also on the monitor's terminal so that the examinee can take notes by hand, even while wearing the non-transparent head mounted display, the examinee can take notes on the gist or keywords in the question if the question is long, and can calculate or think while writing during the answering period. In addition, since the wearer's gaze image data is displayed on the non-transparent head mounted display, the examinee can see outside the non-transparent head mounted display, which creates an environment in which cheating is likely to occur, but since the wearer's gaze image data is also displayed on the monitor's terminal, cheating can be prevented in advance.

<Configuration that detects fraud by providing a light-emitting part to a non-transparent head-mounted display>

In addition, in the above-mentioned test system,
The non-transmissive head mounted display is provided with a light emitting unit,
The test management server
a light emission management means for transmitting a light emission instruction to the examinee terminal via the network for causing the light emitting unit to emit light at any timing during the processing of a question requiring the wearing of a non-transmissive head mounted display, and for receiving, at the timing when the light emission is performed, human image data which is captured by a human image capturing device provided in the examinee terminal and transmitted via the network as light emission confirmation image data, and for executing a process for determining whether or not light emission from the light emitting unit is captured in the received light emission confirmation image data;
The fraud determination means is
It is desirable that the configuration be such that, if the light emission management means is unable to confirm light emission using the light emission confirmation image data, a process is executed to determine that fraud has occurred.

In this way, when a non-transparent head mounted display is provided with a light emitting unit to detect cheating, it is possible to detect cheating such as a person other than the examinee (assistant) wearing the non-transparent head mounted display that should be worn by the examinee and viewing and answering the questions displayed on the display unit, i.e., the assistant viewing and answering questions that should only be visible to the examinee, while the examinee wears a non-transparent head mounted display having the same or a similar shape (another non-transparent HMD that does not display questions) and pretending to take the test, or leaving a photo of an examinee wearing a non-transparent HMD, thereby preventing cheating from occurring. In addition, because the light emitting unit emits light at any time and cannot be predicted on the examinee's terminal, it becomes difficult to return the device to its normal state just before the light is emitted, creating an environment in which cheating is difficult to commit.

<Configuration in which the monitor terminal displays the light-emitting time period in synchronization with the light-emitting part of the non-transparent head-mounted display>

Furthermore, in the case where a light-emitting unit is provided in the non-transmissive head-mounted display to detect fraud,
a monitor terminal connected to the test management server and displaying information for a monitor to monitor the test;
The test management server
a monitoring means for receiving image data including the face of an examinee captured by a person imaging device provided in the examinee terminal and transmitted via a network, and for executing a process of displaying the received image data as a moving image on an observer terminal;
This monitoring means:
It is desirable that a plurality of human image display sections provided on the screen of the monitor terminal display moving image data of each of a plurality of examinees, and that a plurality of light emission time zone display sections provided on the screen of the monitor terminal corresponding to each of the plurality of human image display sections display a light emission time zone indicating that it is a light emission time zone in synchronization with the light emission of the light emission section of the non-transparent head mounted display when a notification that a light emission instruction has been sent from the light emission management means is received.

In this way, if the monitor terminal is configured to display a light-emitting time period in sync with the light emitted by the light-emitting part of the non-transparent head-mounted display to indicate that it is a light-emitting time period, it becomes easier for the monitor to find examinees who may be cheating among the many images of examinees, thereby reducing the burden on the monitor and improving monitoring efficiency.

<Configuration for calculating identity using facial feature score and voiceprint score>

In addition, in the above-mentioned test system,
The test management server
a face image acquisition means for acquiring face image data from person image data including the face of the examinee captured by a person image capture device provided in the examinee terminal and transmitted via a network at the start of the test, during a period during the test when the wearing/detachment status data indicates non-wearing, or at the start of the test, during the period during the test, and at the end of the test;
a facial feature extraction unit that executes a process of extracting facial features from each of the plurality of facial image data acquired by the facial image acquisition unit;
A certificate generating means for executing a process of generating an electronic or paper certificate to be issued to an examinee;
a pre-registration receiving means for receiving voice data for voiceprint extraction, which is acquired at the start of the test by a microphone on the examinee terminal or a microphone on a non-transparent head mounted display and transmitted via a network, performing a voiceprint analysis using the received voice data, and storing the obtained voiceprint data in an examinee information storage means in association with an examinee ID and an examination ID as reference data for calculating identity;
a verification data acquisition means for receiving voice data for voiceprint extraction, which is acquired by a microphone on the examinee terminal or a microphone on a non-transparent head mounted display and transmitted via a network during the test period or at the end of the test in addition to during the test period, performing a voiceprint analysis using the received voice data, and storing the obtained voiceprint data in an examinee information storage means in association with an examinee ID and an examination ID as verification data for calculating identity;
The facial feature extraction means includes:
At the start of the test, the facial feature amount extracted from the facial image data acquired by the facial image acquisition means is stored in the examinee information storage means as reference data for calculating the degree of identity in association with the examinee ID and the test ID;
during the test period or at the end of the test in addition to during the test period, the facial feature amount extracted from the facial image data acquired by the facial image acquisition means is stored in the examinee information storage means as matching data for calculating the identity degree in association with the examinee ID and the test ID, and a process is executed to calculate a facial feature amount score indicating the similarity between the facial feature amount as the matching data for calculating the identity degree and the facial feature amount as the reference data for calculating the identity degree stored in the examinee information storage means,
The face image acquisition means is
At the start of the examination, the facial image data acquired from the person image data is stored in the examinee information storage means in association with the examinee ID and the examination ID as facial image data for a facial photograph to be printed on a certificate to be created by the certificate creation means;
during the period of the test or at the end of the test in addition to the period of the test, only if the facial feature score calculated by the facial feature extraction means is below or equal to a predetermined threshold value, facial image data obtained from the person image data is associated with the examinee ID and the test ID and stored in the examinee information storage means as facial image data for determining fraud, and if the facial feature score is above or equal to the predetermined threshold value, a process is executed not to store the facial image data obtained from the person image data in the examinee information storage means;
The fraud determination means is
calculate a plurality of facial feature scores indicating the similarity between each of the facial feature amounts as matching data for calculating the identity stored in the examinee information storage means and the facial feature amounts as reference data for calculating the identity, or obtain a plurality of facial feature scores calculated by the facial feature extraction means and stored in the examinee information storage means in association with the examinee ID and the test ID, together with the facial feature amounts, and calculate an average facial feature score by averaging the plurality of calculated or obtained facial feature scores;
Furthermore, a plurality of voiceprint scores are calculated which indicate the degree of similarity between each of the voiceprint data as verification data for calculating the degree of identity stored in the examinee information storage means and the voiceprint data as reference data for calculating the degree of identity, and a voiceprint average score is calculated by averaging the plurality of calculated voiceprint scores, and
The method is configured to execute a process of calculating a degree of identity by integrating the calculated average scores of the facial feature amount and the voiceprint average score,
The method for creating the certificate is as follows:
If the identity degree calculated by the fraud determination means is equal to or less than a predetermined threshold value, the certificate may not be produced.

In this way, when the facial feature amount score and voiceprint score are used to calculate the degree of identity, it is possible to more reliably determine whether or not an examinee has been replaced by another person during the test. Furthermore, only when the facial feature amount score is below or equal to a predetermined threshold value (when the similarity of the facial feature amount is low), facial image data is stored in the examinee information storage means, and when the facial feature amount score is equal to or greater than a predetermined threshold value (when the similarity of the facial feature amount is high), the facial feature amount is stored in the examinee information storage means, but the facial image data is not stored in the examinee information storage means, thereby reducing the amount of data to be stored.

<Configuration for issuing instructions to remove non-transparent head-mounted displays and vocal instructions>

Furthermore, in the configuration for calculating the identity using the face feature amount score and the voiceprint score described above,
The test management server
a facial image acquisition necessity determination means for executing a process of transmitting an instruction to remove the non-transmissive head mounted display to the examinee's terminal via the network at any timing in order to obtain facial image data for the time period during the test, when questions requiring the wearing of a non-transmissive head mounted display occur consecutively for a predetermined number of questions or for a predetermined period of time, and when it is determined that the wearing/removal status data during the consecutive questions indicates that the non-transmissive head mounted display is being worn;
a voiceprint acquisition necessity determining means for executing a process of transmitting a voice instruction to an examinee terminal via a network at any timing between questions in order to obtain voice data for voiceprint extraction during the time period during the test, when questions other than those to be answered aloud have appeared in succession for a predetermined number of questions or for a predetermined period of time;
The face image acquisition means is
when the face image acquisition necessity determination means transmits an instruction to remove the non-transmissive head mounted display to the examinee terminal, and when it is determined that the attachment/detachment state data after the removal instruction indicates that the display is not being worn, a process of acquiring face image data from the person image data transmitted from the examinee terminal via the network is executed;
The means for obtaining the matching data is
It is desirable that when a vocal instruction is sent to the examinee's terminal by the voiceprint acquisition necessity determination means, a voiceprint analysis is carried out using the voice data for voiceprint extraction transmitted from the examinee's terminal via the network, and a process is executed to obtain voiceprint data.

When configured in this way to issue instructions to remove the non-transparent head mounted display or to speak, facial feature and voiceprint data can be obtained as matching data for calculating identity even when there are a series of questions during the test that do not require the wearing of the non-transparent head mounted display or questions that are not answered aloud.

<Program invention>

The program of the present invention is intended to cause a computer to function as a test management server that constitutes the test system described above.

The above-mentioned program or a part of it can be recorded on a recording medium such as a magneto-optical disk (MO), a compact disk (CD), a digital versatile disk (DVD), a flexible disk (FD), a magnetic tape, a read-only memory (ROM), an electrically erasable and programmable read-only memory (EEPROM), a flash memory, a random access memory (RAM), a hard disk drive (HDD), a solid-state drive (SSD), a flash disk, etc., and can be stored or distributed. It can also be transmitted using a transmission medium such as a wired network such as a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), the Internet, an intranet, an extranet, or a wireless communication network, or a combination of these, and can also be carried on a carrier wave. Furthermore, the above-mentioned program may be a part of another program, or may be recorded on a recording medium together with a separate program.

As described above, according to the present invention, questions are displayed on a non-transparent head-mounted display, so that no one other than the examinee can see the questions. Also, because the examinee is wearing a non-transparent head-mounted display, the examinee is placed in an environment in which cheating is difficult. Furthermore, because data on the attachment/detachment state from a mounting/detachment detection device provided on the non-transparent head-mounted display is used to determine whether the attachment/detachment state is appropriate, the test can be administered while preventing cheating.

1 is a diagram showing the overall configuration of a test system according to an embodiment of the present invention; FIG. 2 is a perspective view of the non-transmissive head mounted display according to the embodiment. FIG. 2 is a configuration diagram of a non-transmissive head mounted display according to the embodiment. FIG. 4 is a flowchart showing an overall flow of processing related to the test of the embodiment. FIG. 4 is a timing chart of a process relating to the test of the embodiment. FIG. 4 is a diagram showing an example of a screen of an observer terminal according to the embodiment. FIG. 4 is a diagram showing an example of a certificate according to the embodiment. FIG. 13 is a block diagram showing a modified embodiment of the present invention. FIG. 11 is a block diagram showing another modified embodiment of the present invention.

One embodiment of the present invention will be described below with reference to the drawings. Fig. 1 shows the overall configuration of a test system 10 of this embodiment. Figs. 2 and 3 show a perspective view and a configuration diagram of a non-transparent head-mounted display 60. Fig. 4 shows a flow chart of the overall process related to the test, and Fig. 5 shows a timing chart. Fig. 6 shows an example of a screen of an observer terminal 80, and Fig. 7 shows an example of a certificate 300.

<Overall configuration of test system 10>

In FIG. 1, the test system 10 includes a test management server 20, which is connected to a plurality (usually a large number) of examinee terminals 50, one or more monitor terminals 80, one or more evaluator terminals 90, one or more administrator terminals 100, and a plurality (usually a large number) of identity verification requester terminals 110 via a network 1.

Here, network 1 is an external network mainly composed of the Internet, and may be a combination of the Internet and an internal network such as a LAN or an intranet (for example, a limited network constructed within a company or a group of companies), and may be wired or wireless, or may be a combination of wired and wireless. In short, it is sufficient that it can transmit information at a certain speed between multiple points (regardless of distance). Note that the test management server 20 and the monitor terminal 80 do not necessarily need to be connected by network 1, but may be connected by a communication line. Therefore, the monitor terminal 80 may be installed in a remote location from the installation location of the test management server 20, or may be directly connected to the test management server 20.

The test management server 20 is composed of one or more computers, and includes a processing means 20A that executes various processes related to the test, and a question storage means 40, an answer storage means 41, a score storage means 42, an examinee information storage means 43, a test information storage means 44, and a light emission confirmation result storage means 45 that store various data used in the various processes executed by the processing means 20A.

The processing means 20A also includes a pre-registration acceptance means 21, a facial image acquisition means 22, a facial feature extraction means 23, a question data transmission means 24, an answer data reception means 25, a wearing/removing state data reception means 26, an evaluation means 27, a facial image acquisition necessity determination means 28, a score calculation means 29, a certificate creation means 30, a light emission management means 31, a fraud determination means 32, a monitoring means 33, and an identity verification means 34.

Each of the means 21 to 34 included in the processing means 20A is realized by a central processing unit (CPU) provided inside the computer body constituting the test management server 20, one or more programs that define the operating procedures of this CPU, and working memory such as main memory and cache memory. Details of each of these means 21 to 34 will be described later.

Each of the storage means 40 to 45 is preferably realized by, for example, a hard disk drive (HDD) or a solid state drive (SSD), but other recording media such as USB memory, DVD, CD, MO, and magnetic tape may also be used as long as there are no problems with storage capacity or access speed. Details of each of these storage means 40 to 45 will be described later.

Furthermore, when the test management server 20 is configured with multiple computers, various functional distributions can be performed. For example, the processing means 20A can be configured with a WEB server or a WEB application server, and each of the storage means 40-45 can be configured with a database server. In addition, the processing means 20A may be functionally distributed by multiple computers, and each of the storage means 40-45 may be configured with multiple database servers. The data storage format of each of the storage means 40-45 may be a file system rather than a database.

The examinee terminal 50 is a terminal operated by the examinee taking the test, and as shown in FIG. 1, is composed of a main body 50A configured as a computer, a display means 51 such as an LCD display connected to the main body 50A as a peripheral device, an input means 52 such as a keyboard, mouse or touch panel, a person imaging device 53 and an auxiliary imaging device 54 configured as a CCD camera or CMOS camera, a microphone 55, a speaker 56, a fingerprint reader 57 and a non-transparent head mounted display (HMD) 60.

Here, the main body 50A is equipped with a Web browser. A program dedicated to the test may be installed. The main body 50A may be a portable device such as a tablet or a smartphone. The person imaging device 53 is installed to mainly capture images of people (here, examinees) including their faces, like a Web camera used in normal WebRTC (Web Real-Time Communication), but may also capture images of other people including their faces. For example, if the examinee leaves the seat, naturally, images other than people such as background images will be captured. On the other hand, the auxiliary imaging device 54 is installed in one or more units mainly to capture images of objects other than people (examinees) including their faces (for example, hands operating a keyboard, surrounding conditions, etc.), but may be omitted. The auxiliary imaging device 54 may be a device (camera) with the same performance and configuration as the person imaging device 53. The microphone 55 and speaker 56 do not need to be used when the microphone 70 and speaker 71 (see FIG. 3) of the non-transparent head-mounted display 60 are used. Details of the non-transmissive head-mounted display 60 will be described later with reference to Figures 2 and 3.

The monitor terminal 80 is a terminal operated by a monitor who monitors the behavior of each examinee taking the test, and as shown in FIG. 1, is composed of a main body 80A configured by a computer, a display means 81 such as a liquid crystal display connected to the main body 80A as a peripheral device, and an input means 82 such as a keyboard, mouse, or touch panel. The main body 80A is equipped with a web browser. It may also be equipped with a program dedicated to monitoring. The main body 80A may be a portable device such as a tablet.

The evaluator terminal 90 is a terminal operated by an evaluator who evaluates (determines whether the answers are correct or not and assigns marks) the examinees' answers; it is configured with a computer and, although not shown in the figure, includes display means such as an LCD display and input means such as a keyboard, mouse, or touch panel. This evaluator terminal 90 may be a portable device such as a tablet. The examinees' answers may be automatically evaluated by the evaluation means 27 or may be evaluated by an evaluator at the evaluator terminal 90. In the former case of automatic evaluation, the evaluation is usually performed in real time or near real time, but may be performed after the fact. In the latter case of evaluation by an evaluator, the evaluation is performed after the fact.

Specifically, in the latter case, when the evaluator makes a request to the evaluation means 27 of the test management server 20 on the evaluator terminal 90 to obtain the answer data to be evaluated, the answer data stored in the answer storage means 41 is transmitted via the network 1, and the evaluator evaluates the obtained answer data and transmits the evaluation result data to the evaluation means 27 of the test management server 20 via the network 1. In the test management server 20, the evaluation means 27 receives the evaluation result data and stores it in the answer storage means 41 in association with the test ID, examinee ID, and question ID.

The administrator terminal 100 is a terminal operated by the administrator of the test system 10, and is configured with a computer, and includes a display means such as an LCD display and an input means such as a keyboard, mouse, or touch panel, although not shown. This administrator terminal 100 may be a portable device such as a tablet. In the administrator terminal 100, the administrator sets information about the test (test ID, test name, test implementation date, certification organization, certificate form, etc.) and stores it in the test information storage means 44. In addition, the administrator sets information about each question that constitutes each test (test ID, test type, question ID, question type, question time length, answer time length, question data, correct answer data, etc.) and stores it in the question storage means 40.

Here, the exam name is the name of the exam, such as "Xth Y Proficiency Exam (Level 1)" and is used when issuing the certificate. The exam type here does not refer to an English conversation exam, a bar exam, or the like, but rather to information indicating what kind of questions the exam is made up of, such as "only questions requiring HMD wear (an exam consisting of only questions requiring non-transparent HMD wear)", "only questions not requiring HMD wear", "mixed questions requiring and not requiring HMD wear (80% or more required)", or information indicating the strictness, intensity, and severity of the exam (for example, information indicating the type and degree of whether the exam is a high-strictness exam such as a national exam, or an easy-to-take exam such as a test to confirm the effectiveness of daily lessons, and therefore information related to the need to wear a non-transparent HMD). The exam type also includes information indicating whether taking notes during the exam is permitted or not.

Furthermore, the question types here include "HMD required (questions that require wearing a non-transparent HMD)", "HMD not required", "notes allowed (questions that allow taking notes during the test)", "notes not allowed", "vocal answer (questions that require speaking out loud, such as pronouncing a foreign language in a language test)", "multiple choice answer (questions that require silent answering by selecting from options)", etc. The question setting time is the time required to set the question, and for example, in the case of a type of question in which the question is read aloud, it is the length of time required to read it aloud, and in the case of a type of question in which the question and the problem diagram are only displayed and are not read aloud, it is the length of time set by predicting the average time required for the examinee to read the question and understand the contents of the problem diagram. The answering time is the time limit for answering, and can be set to a time that does not allow cheating, such as 20 seconds (for example, a time that is insufficient to cheat and use a dictionary to answer).

The identity verification requester terminal 110 is a terminal operated by an identity verification requester who wishes to verify whether or not the person who presented a certificate presented by another person is the person who was granted the certificate. As shown in FIG. 1, the identity verification requester terminal 110 is configured with a main body 110A formed by a computer, a display means 111 such as a liquid crystal display connected to the main body 110A as a peripheral device, an input means 112 such as a keyboard, mouse, touch panel, etc., an imaging device 113, a microphone 114, and a fingerprint reader 115. The main body 110A is equipped with a web browser. It may also be equipped with a program dedicated to identity verification. The main body 110A may be a portable device such as a tablet or smartphone.

Specifically, when a person who holds a certificate (a person who was a test-taker at the time of the exam) presents his/her certificate at, for example, a job interview at a company, the company that receives the certificate becomes the identity verification requester and can verify the identity of the person by communicating with the identity verification means 34 of the exam management server 20. More details on this will be given later in the explanation of the identity verification means 34 of the exam management server 20.

<Configuration of the test management server 20/pre-registration reception means 21, face image acquisition means 22, and face feature extraction means 23>

In response to a request from the examinee terminal 50, the pre-registration reception means 21 transmits display data for a pre-registration screen (e.g., a web screen) via the network 1, and receives personal information such as the examinee's address, name, date of birth, hobbies, occupation, and student ID number transmitted from the examinee terminal 50 via the network 1, and executes a process of storing the received personal information in the examinee information storage means 43 in association with an automatically assigned examinee ID (examination number, etc.) and test ID. At this time, image data of a photo ID such as a driver's license may also be registered, and this image data can be used for matching processing with face image data acquired by the face image acquisition means 22 described below (matching processing by the fraud determination means 32).

The pre-registration reception means 21 also receives iris data captured by the person imaging device 53 or auxiliary imaging device 54 provided on the examinee terminal 50, or the inward imaging device 69 of the non-transmissive HMD 60, and transmitted via the network 1, and executes a process of storing the received iris data in the examinee information storage means 43 in association with the examinee ID and test ID.

Furthermore, the pre-registration reception means 21 receives voice data for voiceprint extraction acquired by the microphone 55 provided on the examinee terminal 50 or the microphone 70 of the non-transparent HMD 60 and transmitted over the network 1, performs voiceprint analysis using the received voice data, and executes a process of storing the obtained voiceprint data in the examinee information storage means 43 in association with the examinee ID and test ID. At this time, the voiceprint analysis method can be an existing method, and may be a text-dependent method using voice data obtained by vocalizing a fixed phrase, or a text-independent method using voice data of a non-fixed utterance.

The pre-registration reception means 21 also receives fingerprint data that is read by the fingerprint reader 57 provided on the examinee terminal 50 and transmitted via the network 1, and executes a process of storing the received fingerprint data in the examinee information storage means 43 in association with the examinee ID and the test ID.

The face image acquisition means 22 executes a process of receiving person image data (video or still image) captured by the person imaging device 53 provided in the examinee terminal 50 from the examinee terminal 50 via the network 1, detecting and cutting out the face part from the received person image data to acquire face image data, and storing the acquired face image data in the examinee information storage means 43 in association with the examinee ID and the test ID. The face part detection process can adopt an existing method. For example, a feature-based approach by detecting face parts such as eyes and mouth, an image-based approach by pattern recognition such as a neural network or a support vector machine (SVM), or a method combining both of these can be adopted. A method of detecting faces using skin color information may also be adopted. Note that instead of the person image data captured by the person imaging device 53, person image data captured by the inward imaging device 69 of the non-transmissive head mounted display 60 may be used, in which case the inward imaging device 69 becomes the person imaging device. In this case, the non-transparent HMD 60 is not worn, but is held in the hand and the non-transparent HMD 60 is placed at a certain distance from the face so that the entire face is captured.

Furthermore, when the facial image acquisition necessity determination means 28 sends an instruction to remove the non-transparent head mounted display 60 to the examinee terminal 50, and when it is determined that the attachment/detachment state data after the removal instruction indicates that the display is not being worn, the facial image acquisition means 22 also executes a process of acquiring facial image data from the person image data transmitted from the examinee terminal 50 via the network 1. In other words, when the examinee removes the non-transparent HMD 60 in accordance with the instruction to remove the non-transparent HMD 60, the facial image data is acquired from the person image data obtained in the removed state.

The facial feature extraction means 23 executes a process of extracting facial features from the facial image data acquired by the facial image acquisition means 22, and storing the extracted facial features in the examinee information storage means 43 in association with the examinee ID and the test ID. These facial features may be features used in existing face authentication methods executed after face detection, such as the facial contour and the relative positions of facial features such as the eyes, nose, and mouth.

<Configuration of the test management server 20/question data sending means 24, answer data receiving means 25>

The question data transmission means 24 executes a process of transmitting the test question data stored in the question storage means 40 in association with the test ID for the test to be conducted to the test taker terminal 50 via the network 1 in order to display the test question data on the display unit 63 (see Figures 2 and 3) of the non-transparent head-mounted display 60 worn by the test taker.

At this time, the question data may be transmitted to the non-transparent head-mounted display 60 via the main body 50A of the examinee terminal 50 and displayed on the display unit 63, or may be transmitted directly to the non-transparent head-mounted display 60 without passing through the main body 50A and displayed on the display unit 63.

The order in which question data is sent may be a predetermined order common to all examinees, or a random order different for each examinee, or the questions may change depending on the answering status of each examinee (such as the accumulation status of correct/incorrect results in the ongoing test). Candidates taking the same test at the same time will have different speeds at answering the same questions, and the order in which questions are presented or the questions asked may be different. Therefore, the questions that each examinee is processing at a given moment may happen to be the same, but will usually be different.

Furthermore, if necessary, the question data transmission means 24 executes a voice synthesis process to convert the text data, which is the question data, into voice data and transmits it to the examinee terminal 50. Note that the voice synthesis process may be executed on the examinee terminal 50 side.

In addition, during the processing of a question that requires wearing the non-transparent head mounted display 60 and that requires an answer involving a movement, the question data transmission means 24 also executes a process of transmitting, via the network 1 to the examinee terminal 50, a switching instruction to a display mode for action answers in which the wearer's gaze image data captured by the wearer's gaze imaging device 68 (see Figs. 2 and 3) is displayed as a moving image on at least a part of the display unit 63 (see Figs. 2 and 3) of the non-transparent HMD 60 at the timing of transition from the question setting time period to the answer time period (the timing when the question setting time length stored in the question storage means 40 has elapsed from the start of the question setting). Here, displaying the wearer's gaze image data as a moving image on at least a part of the display unit 63 of the non-transparent HMD 60 means that the wearer's gaze image data may be displayed on the entire display unit 63, or that the wearer's gaze image data may be displayed on a part of the display unit 63 while the question text, diagrams, etc. of the question remain on a part of the display unit 63 during the answer time period. In addition, the display unit 63 may display the question text, question diagram, etc. superimposed on the wearer's gaze image data.

The answer data receiving means 25 receives the examinee's answer data transmitted from the examinee terminal 50 via the network 1, and executes a process of storing the received answer data in the answer storage means 41 in association with the test ID, examinee ID, and question ID.

At this time, the examinee inputs the answer using the input means 52 on the examinee terminal 50 according to the question, or inputs the answer by voice using the microphone 55 or the microphone 70 of the non-transparent HMD 60 (see FIG. 3). When voice input is performed, voice recognition processing is performed on the answer data on the answer data receiving means 25 or the examinee terminal 50 side. In addition, operations of the mouse, which is the input means 52 (movement operations, clicking operations, drag-and-drop operations, etc.) may be reflected in operation targets such as a virtual keyboard or buttons displayed on the display unit 63 of the non-transparent HMD 60.

Furthermore, when the display of the display unit 63 of the non-transparent HMD 60 is in the display mode for action answers, the answer data receiving means 25 executes a process of receiving the wearer's gaze image data transmitted from the examinee terminal 50 via the network 1 as answer data. In other words, the wearer's gaze image data becomes the answer data. Specifically, for example, in a question to be answered by writing, the answer data is received as the wearer's gaze image data (video or still image) including the characters and figures written by the examinee. A question to be answered by writing is, for example, a question in which the Japanese word "Good morning" is announced and the examinee writes it in Chinese, or a question in which the Chinese word "Please write a circle" is announced and the examinee writes a circle figure. If the answer is written as a character (including a number or a symbol), the evaluation means 27 executes a character reading process by OCR to convert the answer data into text data, and can perform evaluation (correct/incorrect judgment and grading). If the figure is a handwritten figure, the evaluation means 27 can perform pattern recognition processing using, for example, a neural network, and can evaluate (determine whether the figure is correct or not, and score) the drawn figure. In this case, a model for pattern recognition is constructed in advance using the correct answer data and the incorrect answer data as learning data. That is, a learning process is performed in advance to construct a model for pattern recognition (determine parameters) using the wearer's gaze image data that is determined to be correct and the wearer's gaze image data that is determined to be incorrect, and the corresponding correct and incorrect labels. Although such handwritten input of characters (including numbers and symbols) and figures can also be performed using a digital pen and digital paper, in this case, answer data is obtained from the wearer's gaze image data including characters and figures written on ordinary paper. Therefore, examinees can answer using ordinary writing implements (non-digital writing implements) such as pencils and ballpoint pens and ordinary paper, and can answer in the same way as they would in an ordinary paper test.

The answer data receiving means 25 may receive, as answer data, test tools (real objects) such as red and white blocks, and questions such as "Place the white block on top of the red block," or "Place the red block second from the right out of the five blocks," in a language other than the examinee's native language, in which the examinee actually performs the action, and wearer gaze image data (video) including the action, or wearer gaze image data (still image) including the state after the action is completed. Such actions and the state after the action are completed can be evaluated (correct/incorrect judgment and scoring) by the evaluation means 27, which performs pattern recognition processing using, for example, a neural network. In this case, a model for pattern recognition is constructed in advance using correct answer data and incorrect answer data as learning data. In addition, questions that require the examinee to perform such actions can also be realized by displaying virtual building blocks or the like on the display unit 63 of the non-transparent HMD 60 and having the examinee operate them in a virtual space (VR: virtual reality). In this case, the examinee can answer the questions by performing actions in real space, not in a virtual space, and can answer by actually grabbing and moving test tools (real objects) such as building blocks with his or her own hands, thereby avoiding the hassle of moving objects in a virtual space with a mouse. Depending on the content of the test, the examinee can answer the questions as if they were in a practical test. For example, if the question is about ropework (how to tie a rope) on a ship, the examinee will answer the questions using a rope in real space, so the examinee can answer the questions as if they were in a practical test. Furthermore, a display mode for answering questions by actions in real space using the wearer's gaze image data and a display mode for answering questions by VR operations may coexist, and the examinee may answer in either display mode depending on the test, the question, or the examinee's choice.

<Configuration of the test management server 20/attachment status data receiving means 26>

The wearing/removal state data receiving means 26 receives wearing/removal state data of the non-transparent head mounted display 60 that is detected by the wearing/removal detection device 66 (see Figures 2 and 3) provided on the non-transparent head mounted display 60 and transmitted from the examinee terminal 50 via the network 1, and executes a process of storing the received wearing/removal state data for the question time period and the wearing/removal state data for the answer time period (more specifically, data that combines information indicating whether the question time period or the answer time period is the time period and information indicating whether the display is worn or not) in the answer storage means 41 (the answer storage means 41 in this embodiment also serves as the wearing/removal state storage means) in association with the test ID, examinee ID, and question ID.

In this case, the wearing/removing state data from the wearing/removing detection device 66 of the non-transparent HMD 60 is constantly transmitted from the examinee terminal 50, so the wearing/removing state data receiving means 26 can receive the wearing/removing state data for the question time period and the wearing/removing state data for the answer time period for each question, as shown in FIG. 5. Note that if, during the question time period for a given question, the wearing/removing state data is sometimes data indicating wearing and sometimes data indicating not wearing, the data is stored in the answer storage means 41 as data indicating not wearing overall. Similarly, if, during the answer time period for a given question, the wearing/removing state data is sometimes data indicating wearing and sometimes data indicating not wearing, the data is stored in the answer storage means 41 as data indicating not wearing overall.

<Configuration of test management server 20/evaluation means 27>

The evaluation means 27 evaluates (determines whether the answers are correct or not and scores) the examinee's answer data using the answer data stored in the answer storage means 41 and the correct answer data stored in the question storage means 40, and executes a process of storing the evaluation result data in the answer storage means 41 in association with the test ID, examinee ID, and question ID.

As already mentioned in the explanation of the evaluator terminal 90, the examinee's answers may be automatically evaluated by the evaluation means 27 or may be evaluated by an evaluator at the evaluator terminal 90.

In the former case of automatic evaluation, the evaluation means 27 performs evaluation (correct/incorrect judgment and scoring) by comparing the answer data stored in the answer storage means 41 with the correct answer data stored in the question storage means 40. As already described in the explanation of the answer data receiving means 25, when the wearer's gaze image data is the answer data, the evaluation means 27 performs character reading processing on handwritten input characters (including numbers and symbols) included in the wearer's gaze image data, or performs pattern recognition processing on handwritten input figures included in the wearer's gaze image data, or performs pattern recognition processing on actions and states after completion of actions included in the wearer's gaze image data.

In the latter case of evaluation by an evaluator, the evaluation means 27 responds to a request from the evaluator terminal 90 by transmitting the answer data stored in the answer storage means 41 and the correct answer data stored in the question storage means 40 to the evaluator terminal 90 via the network 1, and receives the evaluation result data by the evaluator transmitted from the evaluator terminal 90 via the network 1, and stores the received evaluation result data in the answer storage means 41 in association with the test ID, examinee ID, and question ID.

<Configuration of the test management server 20/face image acquisition necessity determination means 28>

When questions requiring the wearing of the non-transparent head mounted display 60 occur consecutively for a predetermined number of questions or for a predetermined period of time, and when it is determined that the wearing/removal status data during the consecutive questions indicates that the display is being worn, the facial image acquisition necessity determination means 28 executes a process to send an instruction to remove the non-transparent head mounted display 60 to the examinee terminal 50 via the network 1 at any timing in order to obtain facial image data for the time period during the test, i.e., to obtain facial image data that is different from facial image data before the start of the test and after the end of the test.

As already mentioned in the explanation of the facial image acquisition means 22, when an instruction to remove the non-transparent HMD 60 is sent to the examinee terminal 50, and when the facial image acquisition means 22 determines that the attachment/detachment status data received by the attachment/detachment status data receiving means 26 after the removal instruction is sent is data indicating that the HMD is not being worn, the facial image acquisition means 22 acquires facial image data from the person image data sent from the examinee terminal 50 via the network 1.

<Configuration of the exam management server 20/score calculation means 29, certificate creation means 30>

The grade calculation means 29 uses the evaluation result data stored in the answer storage means 41 to calculate the grades of each examinee, and stores the result of the decision on whether or not to issue a certificate (certificate of passing) in the grade storage means 42 in association with the exam ID, examinee ID, and certificate ID (a certification number, etc., but no number is assigned if a certificate is not issued), or, in the case of an exam in which a certificate is always issued even if the score is poor, executes a process to store the certification level and total score, etc., to be published on the certificate, in association with the exam ID, examinee ID, and certificate ID as grade data in the grade storage means 42. Note that if the cheating determination means 32 determines that cheating has occurred, the grade calculation means 29 does not need to perform the process.

The certificate creation means 30 executes a process of creating an electronic or paper certificate 300 (see FIG. 7) to be issued to the examinee, using the exam name, exam date, certification organization, certificate form, etc. stored in the exam information storage means 44 in association with the exam ID for which the certificate is to be issued, and the score data stored in the score storage means 42. If the cheating determination means 32 determines that cheating has occurred, the certificate will not be issued. Also, if the score does not meet the certification standard, the certificate will not be issued. However, regardless of the score (even if it is poor), a certificate will be issued in the case of an exam that certifies the level (for example, the score in a language test or a grade determined based on the score).

As shown in FIG. 7, certificate 300 lists the certificate title such as "Certificate of Passing" or "Certificate of Completion", the certification ID (certificate number, etc.), name, exam name, certification date (which may be the exam date), certification organization, etc. Certificate 300 also lists a facial photograph 310, a two-dimensional barcode 321 of facial features, a two-dimensional barcode 322 of iris data, a two-dimensional barcode 323 of voiceprint data, a two-dimensional barcode 324 of fingerprint data, and a two-dimensional barcode 325 of personal information such as hobbies.

At this time, if the fraud determination means 32 determines that all of the multiple facial features match or nearly match, the certificate creation means 30 uses the facial image data in which the examinee's face is most clearly shown from the multiple facial image data stored in the examinee information storage means 43 as the facial photograph 310 to be included on the certificate 300, and includes a two-dimensional barcode 321 of the facial features extracted from that facial image data on the certificate 300. The state in which the face is most clearly shown can be appropriately set to, for example, large eye area, facing forward, etc. Each examinee can choose whether or not to include the facial photograph 310.

The certificate creation means 30 also encrypts personal information such as facial features, iris data, voiceprint data, fingerprint data, hobbies, etc., before converting them into a two-dimensional barcode. Each examinee can choose whether or not to include each of the two-dimensional barcodes 321-325. Note that if neither the face photograph 310 nor each of the two-dimensional barcodes 321-325 is included, the probative value of the certificate 300 is reduced.

<Configuration of test management server 20/light emission management means 31>

The light emission management means 31 transmits a light emission instruction to the examinee terminal 50 via the network 1 to cause the light emitting unit 67 (see Figures 2 and 3) to emit light at any timing while processing a question that requires the wearing of the non-transparent head mounted display 60, and receives human image data captured by the human image capture device 53 provided on the examinee terminal 50 and transmitted via the network 1 at the timing when the light emission is performed as light emission confirmation image data, confirms whether or not the light emission of the light emitting unit 67 is captured in the received light emission confirmation image data, and executes a process of storing the light emission confirmation result data in the light emission confirmation result storage means 45 in association with the examinee ID and the test ID.

At this time, the length of time that the light emitting unit 67 of the non-transmissive head mounted display 60 emits light in response to a single light emission command from the light emission management means 31 is not instantaneous, as shown in FIG. 5, but is a length of time that forms a light emission time period of a certain length (for example, one second, several seconds, etc.). Then, during this light emission time period, human image data (image data for light emission confirmation) captured by the human image capture device 53 is transmitted via the network 1. As a result, at least one piece of image data for light emission confirmation is obtained for one light emission command (one light emission time period). If the light emission were instantaneous, when the human image data is transmitted as a video, if the video frame rate (FPS: number of frames per second) is rough, a situation may occur in which the light emission cannot be confirmed even though the light is actually emitted, but by forming a light emission time period, such a situation is avoided.

<Configuration of the test management server 20/fraud determination means 32>

The fraud determination means 32 executes a process to determine whether the attachment/detachment state is appropriate for the question type of the question, using the attachment/detachment state data for the question time period and the attachment/detachment state data for the answer time period received by the attachment/detachment state data receiving means 26 and stored in the answer storage means 41.

Specifically, when the cheating determination means 32 determines that the wearing/removal state data during the question period indicates that the non-transparent head mounted display 60 is being worn and that the wearing/removal state data during the answer period indicates that the non-transparent head mounted display 60 is not being worn (for example, in the case of question (N-2) in FIG. 5), it executes a process of transmitting information to the monitoring means 33 that there is a (slightly high) possibility of cheating (for example, information that a monitoring level requires a warning display using a red frame or the like). Note that the non-transparent HMD 60 may be removed for a short time due to itchy eyes or other reasons, so it is not immediately determined that cheating has occurred when the above-mentioned wearing/removal state is detected.

In addition, when the fraud determination means 32 is processing a question that requires the non-transparent head mounted display 60 to be worn and determines that the wearing/removal status data during the question time period indicates that the display is not being worn (for example, in the case of questions (N-1) and N in FIG. 5), the examinee is in a state where he or she cannot see the question in the first place, and so the fraud determination means 32 may immediately move on to the next question. However, in this embodiment, the fraud determination means 32 continues processing the question as is, and executes a process to transmit information to the monitoring means 33 that there is a (slightly low) possibility of fraud (for example, information that the monitoring level requires a warning display using a yellow frame or the like).

Furthermore, when processing a problem that does not (does not) require the wearing of the non-transparent head mounted display 60, the fraud determination means 32 determines that there is no possibility of fraud unless a determination factor other than the wearing/removal status data indicates the possibility of fraud.

The fraud determination means 32 also performs processing to determine that fraud has occurred if issuance has not been confirmed by processing by the light emission management means 31, using the light emission confirmation result data stored in the light emission confirmation result storage means 45.

Furthermore, the fraud determination means 32 determines whether all of the multiple facial features extracted by the facial feature extraction means 23 match or approximately match, and if all of them do not match or approximately match, it also performs a process to determine that fraud has occurred.

<Configuration of test management server 20/monitoring means 33>

The monitoring means 33 receives image data including the examinee's face captured by the person imaging device 53 provided in the examinee terminal 50 and transmitted via the network 1, and executes a process of displaying the received image data as a video on the monitoring screen 200 of the display means 81 of the monitor terminal 80, as shown in FIG. 6.

In FIG. 6, the monitoring screen 200 has an overall display section 201 that displays images of multiple examinees, and an individual display section 202 that displays images of one or a small number of examinees selected by the monitor.

The overall display unit 201 is provided with a person image display unit 210 that displays moving image data of each of a number of examinees, and a light emission time zone display unit 220 that indicates, by color or the like, that it is a light emission time zone in synchronization with the light emission of the light emitting unit 67 of the non-transmissive HMD 60 when a light emission instruction is issued by the light emission management means 31. The display of the light emission time zone display unit 220 is executed by the monitoring means 33 upon receiving notification that a light emission instruction has been sent from the light emission management means 31.

If the examinee is wearing the non-transparent HMD 60, the wearing state is displayed on the person image display unit 210, so the monitor can check whether the light-emitting unit 67 of the non-transparent HMD 60 is emitting light or not. Since light is not emitted all the time, if the image is constantly emitting light, excluding the case where the light-emitting unit 67 of the non-transparent HMD 60 is broken (the examinee can avoid any disadvantage by proving that it is broken after the fact), it means that cheating such as placing a photo that is emitting light has occurred. On the other hand, if it does not emit light at all, it means that cheating such as wearing a non-transparent HMD other than the non-transparent HMD 60 where the question is displayed, or placing a photo has occurred. Furthermore, if the light emitted by the light-emitting unit 67 of the non-transmissive HMD 60 displayed on the person image display unit 210 is not synchronized with the display of the light-emitting time period display unit 220 by color or other means to indicate that it is a light-emitting time period, this will indicate that fraud is occurring, except in cases where the light-emitting unit 67 of the non-transmissive HMD 60 is broken.

In addition, when the display of the display unit 63 of the non-transparent HMD 60 is switched to a display mode for answering by movement during the processing of a question that requires an answer that involves movement, the person image display unit 210 displays the wearer's gaze image data captured by the wearer's gaze imaging device 68 instead of the person image data.

The individual display unit 202 is provided with a person image display unit 210 that displays the person image data of the examinee selected by the monitor in a moving image, a light emission time zone display unit 220 that indicates the light emission time zone of the examinee, and one or more auxiliary image display units 230 that display the auxiliary image data captured by the auxiliary image capture device 54 in a moving image if one or more auxiliary image capture devices 54 are provided in the examinee terminal 50. As with the overall display unit 201, when the display unit 63 of the non-transparent HMD 60 is switched to a display mode for action answers during the processing of a question that requires an answer involving a movement, the person image display unit 210 of the individual display unit 202 also displays the wearer's gaze image data captured by the wearer's gaze capture device 68 instead of the person image data. In addition, the individual display unit 202 may have both display units for the person image data and the wearer's gaze image data, rather than switching the display. The examinee is selected by the monitor by operating the input means 82, and the monitor means 33 accepts the operation and switches, adds, or deletes the examinees to be displayed. Additionally, the display on the auxiliary image display unit 230 is performed by the monitoring means 33 receiving auxiliary image data transmitted from the examinee terminal 50 via the network 1.

Furthermore, upon receiving a transmission from the fraud determination means 32, the monitoring means 33 executes a warning display process around or in the peripheral area of the examinee's portrait image data displayed on the display means 81 of the monitor terminal 80. As shown in FIG. 6, when the monitoring means 33 receives information from the fraud determination means 32 that there is a (slightly high) possibility of fraud, it executes a warning display process, for example, by surrounding the portrait image display section 210 with a red frame or the like (see examinee number 2 in FIG. 6). Furthermore, when the monitoring means 33 receives information from the fraud determination means 32 that there is a (slightly low) possibility of fraud, it executes a warning display process, for example, by surrounding the portrait image display section 210 with a yellow frame or the like (see examinee number 7 in FIG. 6).

<Configuration of the test management server 20/identification means 34>

The identity verification means 34 uses various identity verification data transmitted from the identity verification requester terminal 110 via the network 1 to execute a process of verifying whether or not the person who presented the certificate 300 (see FIG. 7) to the identity verification requester is the certified person (the person who actually took the test), and executes a process of transmitting the identity verification result data to the identity verification requester terminal 110 via the network 1 and displaying it on the display means 111. For example, if a person who has come to a company for a job interview presents his/her own certificate 300 to the company, the company (the company's interviewer) is the identity verification requester. By referring to the identity verification result displayed on the display means 111 of the identity verification requester terminal 110, the identity verification requester can verify whether or not the person who presented the certificate 300 (the person who has come to the job interview) is the person who took the test related to the presented certificate 300.

Specifically, first, if the presented certificate 300 includes a facial photograph 310, the person making the identity verification request visually compares the face of the person presenting the certificate 300 (the person who has come for the job interview) with the facial photograph 310 on the certificate 300 to determine whether they are the same person.

In addition, if the presented certificate 300 contains a two-dimensional barcode 321 of facial features, the person requesting identity verification uses the imaging device 113 on the identity verification requester terminal 110 to capture personal image data including the face of the person who has come to the job interview, and transmits the captured personal image data to the identity verification means 34 of the test management server 20 via network 1, and also uses the imaging device 113 to capture the two-dimensional barcode 321 of the facial features of the presented certificate 300, and transmits the captured two-dimensional barcode 321 of the facial features (in an encrypted state) to the identity verification means 34 of the test management server 20 via network 1. In the test management server 20, the identity verification means 34 receives current person image data transmitted from the identity verification requester terminal 110 via network 1, obtains facial image data from the received person image data, and further extracts current facial features from the facial image data. It also receives a two-dimensional barcode 321 of the facial features transmitted from the identity verification requester terminal 110 via network 1, performs a decoding process to obtain the facial features at the time of the test, determines whether the facial features match or nearly match, and transmits the identity verification result data to the identity verification requester terminal 110 via network 1.

Furthermore, if the presented certificate 300 contains a two-dimensional barcode 322 of iris data, the person requesting identity verification uses the imaging device 113 on the identity verification requester terminal 110 to capture the iris data of the person who has come to the job interview and transmits the captured iris data to the identity verification means 34 of the test management server 20 via network 1, and also uses the imaging device 113 to capture the two-dimensional barcode 322 of the iris data of the presented certificate 300 and transmits the two-dimensional barcode 322 of the captured iris data (in an encrypted state) to the identity verification means 34 of the test management server 20 via network 1. Then, in the test management server 20, the identity verification means 34 receives the current iris data transmitted from the identity verification requester terminal 110 via network 1, and also receives the two-dimensional barcode 322 of the iris data transmitted from the identity verification requester terminal 110 via network 1, performs a decoding process to obtain the iris data at the time of the test, determines whether the two sets of iris data match or nearly match, and transmits the identity verification result data to the identity verification requester terminal 110 via network 1.

In addition, if the presented certificate 300 contains a two-dimensional barcode 323 of voiceprint data, the person requesting identity verification acquires voice data of the person who comes to the job interview using the microphone 114 on the identity verification requester terminal 110 and transmits the acquired voice data to the identity verification means 34 of the test management server 20 via network 1, and also captures an image of the two-dimensional barcode 323 of the voiceprint data on the presented certificate 300 using the imaging device 113 and transmits the captured two-dimensional barcode 323 of the voiceprint data (in an encrypted state) to the identity verification means 34 of the test management server 20 via network 1. In the test management server 20, the identity verification means 34 receives the voice data transmitted from the identity verification requester terminal 110 via network 1, performs a voiceprint analysis using the received voice data (the same method as the voiceprint analysis performed by the pre-registration reception means 21) to obtain current voiceprint data, and receives the two-dimensional barcode 323 of the voiceprint data transmitted from the identity verification requester terminal 110 via network 1, performs a decryption process to obtain the voiceprint data at the time of the test, determines whether the two voiceprint data match or nearly match, and transmits the identity verification result data to the identity verification requester terminal 110 via network 1.

Furthermore, if the presented certificate 300 contains a two-dimensional barcode 324 of fingerprint data, the person requesting identity verification acquires fingerprint data of the person who has come to the job interview using the fingerprint reading device 115 on the identity verification requester terminal 110, and transmits the acquired fingerprint data to the identity verification means 34 of the test management server 20 via network 1, and also captures an image of the two-dimensional barcode 324 of the fingerprint data on the presented certificate 300 using the imaging device 113, and transmits the captured two-dimensional barcode 324 of the fingerprint data (in an encrypted state) to the identity verification means 34 of the test management server 20 via network 1. Then, in the test management server 20, the identity verification means 34 receives the current fingerprint data transmitted from the identity verification requester terminal 110 via network 1, and also receives the two-dimensional barcode 324 of the fingerprint data transmitted from the identity verification requester terminal 110 via network 1, performs a decryption process to obtain the fingerprint data at the time of the test, determines whether the two sets of fingerprint data match or nearly match, and transmits the identity verification result data to the identity verification requester terminal 110 via network 1.

In addition, if the presented certificate 300 contains a two-dimensional barcode 325 of personal information, the person requesting identity verification asks the person who has come to the job interview personal information such as name, date of birth, hobbies, etc., and operates the input means 112 on the identity verification requester terminal 110 to input the obtained personal information, which is then transmitted to the identity verification means 34 of the test management server 20 via network 1, and the imaging device 113 captures an image of the two-dimensional barcode 325 of the personal information on the presented certificate 300, and transmits the captured two-dimensional barcode 325 of the personal information (in an encrypted state) to the identity verification means 34 of the test management server 20 via network 1. In the test management server 20, the identity verification means 34 receives the current personal information sent from the identity verification requester terminal 110 via network 1, and also receives the two-dimensional barcode 325 of the personal information sent from the identity verification requester terminal 110 via network 1, performs a decryption process to obtain the personal information at the time of the test, determines whether the two sets of personal information match or nearly match, and sends the identity verification result data to the identity verification requester terminal 110 via network 1.

<Configuration of the test management server 20/question storage means 40, answer storage means 41, score storage means 42, examinee information storage means 43, test information storage means 44, light emission confirmation result storage means 45>

As shown in FIG. 4, the question storage means 40 stores the test ID, test type, question ID, question type, question time length, answer time length, question data, and correct answer data in association with each other.

As shown in FIG. 4, the answer storage means 41 stores the test ID, examinee ID, question ID, answer data, wearing/removing state data for the question time period, wearing/removing state data for the answer time period, and evaluation result data in association with each other. Note that the answer storage means 41 in this embodiment also serves as the wearing/removing state storage means. If a wearing/removing state storage means is provided separately, the test ID, examinee ID, question ID, wearing/removing state data for the question time period, and wearing/removing state data for the answer time period are stored in association with each other.

As shown in FIG. 4, the score storage means 42 stores the test ID, examinee ID, score data, and certificate ID in association with each other.

As shown in FIG. 4, the examinee information storage means 43 stores examinee ID, test ID, address, name, date of birth, hobbies, occupation, face image data (multiple data including candidates), face feature values (multiple data including candidates), iris data, voiceprint data, fingerprint data, etc. in association with each other.

As shown in FIG. 4, the test information storage means 44 stores test ID, test name, test implementation date, certification organization, certificate form, etc. in association with each other.

The light emission confirmation result storage means 45 stores the examinee ID, test ID, and light emission confirmation result data (multiple pieces) in association with each other, as shown in FIG. 4.

<Configuration of non-transparent head-mounted display 60: Figures 2 and 3>

2 and 3, the non-transparent head mounted display 60 comprises a main body 60A, a first belt 61 (a belt that contacts the top of the head) and a second belt 62 (a belt that goes around the left and right sides of the head and the back of the head) for fixing the main body 60A to the head of the wearer, a display unit 63 embedded in the main body 60A, a control unit 64 that performs various controls such as input/output control including display and communication control, a communication unit 65 that communicates with the main body 50A of the examinee terminal 50 or the test management server 20, and a non-transparent It is equipped with an attachment/detachment detection device 66 that detects the attachment/detachment state of the non-transmissive head mounted display 60, a light emitting unit 67 that emits light in response to a light emission instruction from the light emission management means 31, a wearer's line of sight imaging device 68 that can image the space outside the non-transmissive head mounted display 60 from the line of sight of the wearer of the non-transmissive head mounted display 60, an inward direction imaging device 69 that can image the inward direction of the non-transmissive head mounted display 60 (the direction of the wearer's face), a microphone 70, and a speaker 71.

In this embodiment, the attachment/detachment detection device 66 is provided, as an example, at least one of the connection parts 61A of the first belt 61 and the connection parts 62A and 62B of the second belt 62. In the case of three attachment/detachment detection devices 66A, 66B, and 66C, when all the attachment/detachment detection devices are in an attached state, data indicating that the devices are attached as a whole is output, and when any of the attachment/detachment detection devices is in an unattached state, data indicating that the devices are not attached as a whole is output, so that it may be considered that one attachment/detachment detection device 66 is configured as a whole. This attachment/detachment detection device 66 may be configured to detect the attachment/detachment state by detecting the contact/non-contact of the hooks that constitute the connection part, may be configured to detect the attachment/detachment state by capturing the interruption of the laser light when one of the connection parts is inserted into the other, may be configured to detect the attachment/detachment state by detecting the gap between one of the connection parts and the other with a gap sensor, and may be configured to detect the attachment/detachment state by any method of attachment/detachment detection. In this embodiment, the attachment/detachment detection device 66 is provided at a connection partway along the belt, but it may also be provided at the connection part between the belt and the main body 60A, and does not necessarily have to be provided at the connection part. For example, the attachment/detachment detection method may be a method in which the device is provided on the main body 60A and measures the distance between the main body 60A and the wearer's head.

The wearer's line of sight imaging device 68 and the inward direction imaging device 69 are composed of, for example, a CCD camera or a CMOS camera. In addition, in tests and questions that do not require the wearing of the non-transmissive head mounted display 60, both the wearer's line of sight imaging device 68 and the inward direction imaging device 69 can be used as a person imaging device 53 for mainly capturing images of a person (here, the examinee) including a face, and can also be used as an auxiliary imaging device 54 for capturing images of objects other than a person (the examinee) including a face (for example, a hand operating a keyboard, surrounding conditions, etc.). In this case, the imaging direction (direction of the center line of the camera) of the wearer's line of sight imaging device 68 and the inward direction imaging device 69 needs to be fixed in a state facing a person including a face or other objects (hands operating a keyboard, etc.) depending on the purpose, so it is preferable to prepare a dedicated support for fixing the main body 60A of the non-transmissive HMD 60 in a desired position, or to make the outer shape of the main body 60A of the non-transmissive HMD 60 into a shape suitable for fixing. The former dedicated support is, for example, a support such as a camera tripod. The latter shape suitable for fixing is, for example, when the bottom part of the main body 60A is formed so that the imaging direction of the wearer's line of sight imaging device 68 and the inward direction imaging device 69 faces the desired direction when the non-transparent HMD 60 is placed on a desk, or when the bottom part of the main body 60A is provided with a claw-shaped part (similar to the claw-shaped part for fixing provided on a normal Web camera) that can be attached by hooking it onto the upper end of the display means 51, etc.

The display of the display unit 63 has a normal display mode, a display mode for action answers, and a display mode for notes. In the normal display mode, the problem statement, problem diagrams, etc. are displayed on the entire screen of the display unit 63.

In the display mode for action-based answers, the question requires wearing the non-transmissive head mounted display 60, and during the processing of a question that requires an answer involving a movement, at the timing of the transition from the question time period to the answer time period, the wearer's gaze image data captured by the wearer's gaze capturing device 68 is displayed as a moving image on at least a part of the display unit 63 of the non-transmissive HMD 60. The details of this display mode for action-based answers have already been described in detail in the explanation of the answer data receiving means 25, so a detailed explanation will be omitted here.

In the display mode for notes, during the processing of a question that requires wearing the non-transparent head mounted display 60 and allows written notes during the question and answer period, the test question data is displayed on the display unit 63 of the non-transparent HMD 60, and the wearer's gaze image data captured by the wearer's gaze imaging device 68 is displayed so that the examinee can take notes by hand. In this case, the display area of the test question data on the display unit 63 and the display area of the wearer's gaze image data may be separated, or the two data may be superimposed. When both display areas are separated as in the former case, for example, they may be separated into left and right or top and bottom, and one of them may be used as a memo area (display area of the wearer's gaze image data), or they may be separated into the upper right part and the rest, the lower right part and the rest, the upper left part and the rest, or the lower left part and the rest, and one of them may be used as a memo area. This allows the examinee to take notes on the gist or keywords in the question text if it is a long-form question during the question period. In other words, you can write notes on paper (not digital paper, but regular paper) using a writing implement such as a pencil or ballpoint pen with your own hand. During the answering period, you can also calculate or think while writing.

The display unit 63 may be configured by fitting a smartphone into the main body 60A of the non-transparent HMD 60. In this configuration, if a light-emitting unit 67 is provided, the light-emitting unit 67 must be emitted from the smartphone that constitutes the display unit 63. This is because the main body 60A on which the light-emitting unit 67 is provided and the smartphone fitted into the main body 60A are not integrated and can be separated, so that a person other than the examinee (assistant) can remove the smartphone from the main body 60A and understand the questions by looking at the smartphone's display screen while hiding in another place. Therefore, the smartphone and the main body 60A are connected by a cable or the like, and the light-emitting unit 67 is made to emit light by a signal from the smartphone, or a part of the light (light to form the light-emitting unit 67) of the smartphone's display screen (the surface facing the wearer's face) is refracted using a mirror or the like to be directed in the opposite direction.

The microphone 70 and speaker 71 may be configured by connecting a headset separate from the main body 60A of the non-transparent HMD 60 to the main body 60A (not shown in FIG. 2), or may be built into the main body 60A. Note that when using the microphone 55 and speaker 56 connected to the main body 50A of the examinee terminal 50, the installation of the microphone 70 and speaker 71 may be omitted.

<Overall flow of testing processes: Figure 4>

In this embodiment, various tests are performed by the test system 10 according to the following process flow:

In FIG. 4, first, the examinee operates the examinee terminal 50, inputs personal information such as his/her address, name, date of birth, hobbies, and occupation as pre-registration data, and transmits it to the examination management server 20 via the network 1 (step S1). The examination management server 20 receives the personal information such as the examinee's hobbies transmitted from the examinee terminal 50 via the network 1 by the pre-registration reception means 21, and stores the received personal information in the examinee information storage means 43 in association with the automatically assigned examinee ID (examination number, etc.) and examination ID (step S2). At this time, the examinee may also transmit image data, iris data, voiceprint data, and fingerprint data of a photo ID such as a driver's license, if necessary, and store them in the examinee information storage means 43. The details of these processes have already been described in detail in the description of the pre-registration reception means 21, so a detailed description will be omitted here.

Next, the examinee operates the examinee terminal 50, captures image data including his/her face using the person image capture device 53 in order to register the facial image data and facial feature values before the start of the test, and transmits the captured image data to the test management server 20 via the network 1 (step S3). In the test management server 20, the facial image acquisition means 22 receives the facial image data transmitted from the examinee terminal 50 via the network 1, detects and cuts out the facial portion from the received facial image data to obtain facial image data, associates the obtained facial image data with the examinee ID and the test ID and stores it in the examinee information storage means 43, and the facial feature value extraction means 23 extracts facial feature values from the facial image data, associates the extracted facial feature values with the examinee ID and the test ID and stores them in the examinee information storage means 43 (step S4). The details of these processes have already been described in detail in the description of the facial image acquisition means 22 and the facial feature value extraction means 23, so a detailed description will be omitted here.

Then, the test begins. Here, we will explain the case of a test that requires wearing a non-transparent head mounted display 60. At the examinee terminal 50, the examinee wears the non-transparent HMD 60 to prepare for the test. Then, in the test management server 20, the question data transmission means 24 transmits the test question data stored in the question storage means 40 in association with the test ID for the test to be conducted to the examinee terminal 50 via the network 1 in order to display the test question data on the display unit 63 (see Figures 2 and 3) of the non-transparent HMD 60 worn by the examinee (step S5).

The examinee terminal 50 receives the question data transmitted from the test management server 20 over the network 1 via the non-transparent head mounted display 60, either via the main body 50A or directly, and starts playing back the question data (question statements, question diagrams, etc.) (step S6). This playback is primarily displayed on the display unit 63, but may also include audio output. After playback starts, the presentation of the questions (questioning) continues for the question duration stored in the question storage means 40, and this period constitutes the questioning time period.

During this question period, the attachment/detachment status data output from the attachment/detachment detection device 66 of the non-transparent head mounted display 60 is transmitted to the test management server 20 via the network 1 (step S7). This transmission may be via the main body 50A of the examinee terminal 50, or may be sent directly from the non-transparent HMD 60 to the network 1 without passing through the main body 50A.

In the test management server 20, the wearing/removing state data receiving means 26 receives the wearing/removing state data transmitted from the non-transparent HMD 60 via the network 1, and stores the received wearing/removing state data for the question time period in the answer storage means 41 in association with the test ID, examinee ID, and question ID (step S8).

The fraud determination means 32 also uses the attachment/detachment state data for the time period in which the question was asked that is stored in the answer storage means 41 to determine whether the attachment/detachment state is appropriate for the type of question asked (determines whether a warning display is required), and if the condition is met, transmits information that there is a possibility of fraud to the monitoring means 33. In response to this transmission, the monitoring means 33 displays a warning on the monitor terminal 80 (step S8). The details of these processes have already been described in detail in the explanation of the fraud determination means 32 and the monitoring means 33, so a detailed explanation will be omitted here.

For questions that require an answer that involves a movement, when the question period ends and the answer period begins, the question data transmission means 24 transmits an instruction to switch to the display mode for action answers to the examinee terminal 50 via the network 1 (step S9). When the non-transparent head-mounted display 60 of the examinee terminal 50 receives the instruction to switch to the display mode for action answers via the main body 50A or directly, the wearer's gaze image data captured by the wearer's gaze imaging device 68 (see Figures 2 and 3) is displayed as a moving image on at least a part of the display unit 63 of the non-transparent HMD 60 (step S10).

Then, even during the answering period, the attachment/detachment status data output from the attachment/detachment detection device 66 of the non-transparent head-mounted display 60 is transmitted to the test management server 20 via the network 1 (step S11).

In the test management server 20, the wearing/removing state data receiving means 26 receives the wearing/removing state data transmitted from the non-transparent HMD 60 via the network 1, and stores the received wearing/removing state data for the answer time period in the answer storage means 41 in association with the test ID, examinee ID, and question ID (step S12).

The fraud determination means 32 also uses the attachment/detachment state data for the question time period and the attachment/detachment state data for the answer time period stored in the answer storage means 41 (that is, since the answer time period has begun, the attachment/detachment state data for the previous question time period is also used) to determine whether the attachment/detachment state is appropriate for the question type of the question asked (determines whether a warning display is required), and if the conditions are met, transmits information to the monitoring means 33 that there is a possibility of fraud. Upon receiving this transmission, the monitoring means 33 displays a warning on the monitor terminal 80 (step S12).

Furthermore, the light emission management means 31 determines whether or not to emit light from the light emitting unit 67 (see Figures 2 and 3) of the non-transparent HMD 60 (light emission is performed at any timing), and if it is to be emitted, transmits a light emission instruction to the examinee terminal 50 via the network 1 (step S13). The examinee terminal 50 receives the light emission instruction at the non-transparent HMD 60 via the main body 50A or directly, emits light from the light emitting unit 67, and transmits the person image data captured by the person image capture device 53 at the timing when the light emission was performed as light emission confirmation image data to the test management server 20 via the network 1 (step S14). The test management server 20 receives the light emission confirmation image data transmitted via the network 1 by the light emission management means 31, checks whether the light emission from the light emitting unit 67 is reflected in the received light emission confirmation image data, and stores the light emission confirmation result data in the light emission confirmation result storage means 45 in association with the examinee ID and the test ID (step S13). The details of these processes have already been described in detail in the explanation of the light emission management means 31, so a detailed explanation will be omitted here. Note that, for the sake of convenience, in steps S13 and S14, it is described that light is emitted during the answer time period, but light may be emitted during the question setting time period, the answer time period, or it may span both time periods, and in this embodiment, light emission management is performed at a timing separate from the setting and answering of each question.

Next, the examinee inputs the answer to the question on the examinee terminal 50 and transmits the answer data to the exam management server 20 via the network 1 (step S15). At this time, for questions that require an answer involving movement, the examinee answers while looking at the wearer's gaze image data showing his or her own movements on the display unit 63 of the non-transparent HMD 60. In the exam management server 20, the answer data receiving means 25 receives the examinee's answer data transmitted from the examinee terminal 50 via the network 1, and stores the received answer data in the answer storage means 41 in association with the exam ID, examinee ID, and question ID (step S16). The details of these processes have already been described in detail in the explanation of the answer data receiving means 25, so a detailed explanation will be omitted here.

Then, in the test management server 20, the evaluation means 27 evaluates (determines whether the answer is correct or not and scores) the examinee's answer data using the answer data stored in the answer storage means 41 and the correct answer data stored in the question storage means 40, and stores the evaluation result data in the answer storage means 41 in association with the test ID, examinee ID, and question ID (step S17). The details of these processes have already been described in detail in the explanation of the evaluation means 27, so a detailed explanation will be omitted here.

Next, in the test management server 20, if the facial image acquisition necessity determination means 28 determines that questions requiring the wearing of the non-transparent head mounted display 60 occur consecutively for a predetermined number of questions or for a predetermined period of time, and that the wearing/removal status data during the consecutive questions indicates that the display is being worn, it determines whether or not to acquire facial image data for the time period during the test (i.e., facial image data that is different from facial image data before the start of the test or after the end of the test) (the acquisition of facial image data for the time period during the test can be performed at any time), and if so, sends an instruction to remove the non-transparent HMD 60 to the examinee terminal 50 via the network 1 (step S18).

The examinee terminal 50 receives the instruction to remove the non-transparent HMD 60 and displays the instruction on the display unit 63 (step S19). The removal instruction may be given by voice output. Then, the examinee removes the non-transparent HMD 60 in accordance with the removal instruction, and then uses the person image capture device 53 to capture image data including his/her face, and transmits the captured image data to the test management server 20 via the network 1 (step S20). In the test management server 20, the face image capture means 22 receives the image data transmitted from the examinee terminal 50 via the network 1, detects and cuts out the face portion from the received image data to obtain face image data, associates the obtained face image data with the examinee ID and the test ID, and stores the acquired face image data in the examinee information storage means 43, and uses the face feature extraction means 23 to extract face features from the face image data, and associates the extracted face features with the examinee ID and the test ID and stores them in the examinee information storage means 43 (step S21). This process is the same as the process in step S4 described above.

Then, it is determined whether or not to proceed to the next question (step S22). If it is determined that the test should proceed to the next question, the process returns to step S5 described above, and thereafter steps S5 to S22 are repeated until it is determined that the test should not proceed to the next question. On the other hand, if the test should not proceed to the next question, the test ends.

After the test is completed, the examinee removes the non-transparent HMD 60 and captures image data including his/her face using the person image capture device 53, and transmits the captured image data to the test management server 20 via the network 1 (step S23). In the test management server 20, the face image capture means 22 receives the image data transmitted from the examinee terminal 50 via the network 1, obtains face image data by detecting and cutting out the face from the received image data, and stores the obtained face image data in the examinee information storage means 43 in association with the examinee ID and the test ID, while the face feature extraction means 23 extracts face features from the face image data, and stores the extracted face features in the examinee information storage means 43 in association with the examinee ID and the test ID (step S24). This process is the same as the processes in steps S4 and S21 described above.

Next, in the exam management server 20, the score calculation means 29 uses the evaluation result data stored in the answer storage means 41 to calculate the score of each examinee, and stores the result of the decision on whether to issue a certificate 300 (see FIG. 7) in the score storage means 42 in association with the exam ID, examinee ID, and certificate ID (a certification number, etc., but no number is assigned if a certificate is not issued), as score data, or in the case of an exam in which a certificate is always issued even if the score is poor, the certification level and total score to be published on the certificate are stored in association with the exam ID, examinee ID, and certificate ID as score data in the score storage means 42 (step S25). At this time, the cheating determination means 32 also performs a process to determine whether cheating has occurred, and if it is determined that cheating has occurred, the score calculation means 29 does not need to perform the process.

Then, the certificate creation means 30 creates an electronic or paper certificate 300 (see Figure 7) to be issued to each examinee using the exam name, exam date, certification organization, certificate form, etc. stored in the exam information storage means 44 in association with the exam ID for which the certificate is to be issued, and the score data of each examinee stored in the score storage means 42 (step S26).

<Effects of this embodiment>

This embodiment has the following advantages. In other words, in the examination system 10, questions are displayed on the non-transparent head-mounted display 60, so that no one other than the examinee can see the questions. In addition, because the examinee is wearing the non-transparent HMD 60, it is possible to create an environment in which it is difficult to commit cheating, such as secretly looking at a dictionary.

Furthermore, the fraud determination means 32 uses the attachment/detachment state data detected by the attachment/detachment detection device 66 of the non-transparent head mounted display 60 to determine whether the attachment state is appropriate for the type of question posed, so that it can detect, for example, a situation where a test subject answers a question that requires the non-transparent HMD 60 without wearing the non-transparent HMD 60, or a situation where the non-transparent HMD 60 is not being worn when the test subject should be wearing the non-transparent HMD 60 to view the question. This makes it possible to conduct the test while preventing fraud.

In addition, since the examination system 10 is equipped with a monitoring means 33 and an observer terminal 80, a warning message can be displayed on the observer terminal 80 (see FIG. 6). This allows an observer who is monitoring multiple examinees at once to be notified of an examinee who may be cheating, reducing the observer's monitoring burden and improving monitoring efficiency.

Furthermore, the examination system 10 includes a facial image acquisition means 22, a facial feature extraction means 23, a fraud determination means 32, and a certificate creation means 30, and is therefore able to extract facial features for each of a plurality of facial image data, determine whether all of the plurality of facial feature data match or approximately match, and, if it is determined that all match or approximately match, use the facial image data among the plurality of facial image data which most clearly shows the examinee's face as the facial photograph 310 to be included on the certificate 300 (see FIG. 7). This makes it possible to determine whether or not people have been replaced during the examination and detect fraud, as well as to obtain an appropriate facial photograph 310 to be included on the certificate 300.

The examination system 10 is also equipped with a facial image acquisition necessity determination means 28, and is therefore able to issue an instruction to remove the non-transparent head mounted display 60. This means that even in situations where it is difficult to obtain facial image data during the examination due to a series of questions requiring the wearing of the non-transparent HMD 60, the necessary facial image data can be obtained and cheating can be detected. In addition, the instruction to remove the non-transparent HMD 60 is sent at any timing by the facial image acquisition necessity determination means 28 when certain conditions are met, and therefore cannot be predicted on the examinee terminal 50 side, and therefore, for example, it will be impossible for the examinee to sense that a removal instruction has been issued and to take action such as returning from the assistant to the examinee, thereby creating an environment in which cheating is difficult.

In addition, in the test system 10, the non-transparent HMD 60 is provided with a wearer's line of sight imaging device 68, so that in questions that require an answer involving movement, the wearer's line of sight image data captured by the wearer's line of sight imaging device 68 can be displayed on the display unit 63 of the non-transparent HMD 60 and can also be displayed on the monitor terminal 80. Therefore, when a question that requires an answer involving movement is asked, even while wearing the non-transparent HMD 60, the examinee can answer the question while checking the movement of his or her hands, fingers, etc. with his or her own eyes. In addition, since the wearer's line of sight image data is displayed on the non-transparent HMD 60, the examinee can see the outside of the non-transparent HMD 60, which creates an environment in which cheating is more likely to occur, but since the wearer's line of sight image data is also displayed on the monitor terminal 80, cheating can be prevented in advance.

When the wearer's gaze image data captured by the wearer's gaze imaging device 68 as described above is displayed on the display unit 63 of the non-transmissive HMD 60, and the examinee writes an answer, the evaluation means 27 can execute a character reading process on the wearer's gaze image data to evaluate the answer (such as determining whether it is correct or not and marking it). This makes it possible to automate the evaluation.

In addition, in the test system 10, the non-transparent HMD 60 is provided with a wearer's gaze imaging device 68, so that for questions for which taking notes is permitted, the wearer's gaze image data can be displayed on the display unit 63 of the non-transparent HMD 60 so that the examinee can take notes by hand, and can also be displayed on the monitor terminal 80. Therefore, even if the non-transparent HMD 60 is still worn, the examinee can take notes on the gist or keywords in the question text if the question is long, during the question period, and can calculate or think while writing during the answer period. In addition, since the wearer's gaze image data is displayed on the display unit 63 of the non-transparent HMD 60, the examinee can see the outside of the non-transparent HMD 60, which creates an environment in which cheating is easy to commit, but since the wearer's gaze image data is also displayed on the monitor terminal 80, cheating can be prevented in advance.

Furthermore, the test system 10 includes a light emission management means 41, and the non-transparent head mounted display 60 is provided with a light emitting unit 67. This allows a person other than the examinee (assistant) to wear the non-transparent HMD 60 that should be worn by the examinee, and answer questions displayed on the display unit 63 by looking at them. In other words, the assistant looks at and answers questions that are normally only visible to the examinee, while the examinee wears a non-transparent HMD having the same or similar shape (another non-transparent HMD that does not display questions) and pretends to take the test. It is also possible to detect cheating such as leaving a photo of the examinee wearing the non-transparent HMD 60. This makes it possible to prevent cheating. In addition, since the light emitting unit 67 emits light at any timing, it cannot be predicted on the examinee terminal 50 side, so it is difficult to return to a normal state just before the light is emitted, and an environment in which cheating is difficult to perform can be created.

<Transformation forms>

The present invention is not limited to the above-described embodiment, and any modifications that can achieve the object of the present invention are included in the present invention.

<The examinee terminal 50 is configured with only a non-transparent HMD 60, and a smartphone is fitted into the main body 60A of the non-transparent HMD 60 to form the display unit 63>

For example, in the above embodiment, the examinee terminal 50 includes a main body 50A configured by a computer and a non-transparent HMD 60, but the examinee terminal may be configured only with the non-transparent HMD 60 without using the main body 50A configured by a computer. Also, when the examinee terminal is configured only with the non-transparent HMD 60 without using the main body 50A configured by a computer, a smartphone may be fitted into the main body 60A of the non-transparent HMD 60 to form the display unit 63. In such a configuration, there is no human image capture device 53 that captures an image of an examinee wearing the non-transparent HMD 60 from the outside, so the observer cannot monitor the examinee using the observer terminal 80, and it is also not possible to obtain light emission confirmation image data for confirming the light emission of the light emitting unit 67 of the non-transparent HMD 60, but it is possible to prevent fraud using the wearing/removing state data.

<Configuration using an imaging device installed on a mobile device such as a smartphone or tablet>

In addition, in the above embodiment, the person imaging device 53 and the auxiliary imaging device 54 are connected to the main body 50A of the examinee terminal 50 (see FIG. 1), but these person imaging devices 53 and the auxiliary imaging devices 54 may be configured as imaging devices (cameras) installed on a mobile device such as a smartphone or tablet terminal. In this case, image data captured by the mobile device may be transmitted from the mobile device to the test management server 20 via the main body 50A, or may be transmitted directly from the mobile device to the test management server 20 without passing through the main body 50A. Note that the main body 50A may be configured as a mobile device such as a smartphone or tablet terminal, and in that case, while displaying questions on the screen of the first mobile device (e.g., a tablet terminal), imaging processing of a person image including the examinee's face, an image of the hand, etc. may be performed using imaging devices (cameras) of the second, third, ... mobile devices (e.g., a smartphone or another tablet terminal).

<Configuration for advertising and selling study materials useful for exam preparation, and utilizing data on study material purchasers (test takers): Figure 8>

As shown in FIG. 8, the embodiment may be configured to advertise and sell educational materials useful for exam study, and to utilize data on purchasers (examinees) of the educational materials. In such a configuration, although not shown in FIG. 8, all of the functions provided in the embodiment may be provided. Therefore, FIG. 8 is mainly described with respect to functions added to the configuration of the embodiment. That is, the processing means 20A of the exam management server 20 includes a educational material purchase acceptance means 35, a educational material usage result aggregation means 36, an advertisement data creation means 37, and an advertisement posting means 38. The exam management server 20 also includes a educational material usage status storage means 46, a educational material usage result storage means 47, and an advertisement data storage means 48.

The teaching material purchase acceptance means 35 accepts an application for purchasing teaching materials from an examinee who clicked on an advertisement for teaching materials displayed on the display unit 63 of the non-transparent head-mounted display 60 after the end of the test (after the processing of step S25 in FIG. 4 in the above embodiment, etc.), and executes a process of storing in the teaching material usage status storage means 46 the name, date of birth, and other personal identification information of the examinee who purchased the teaching materials (address, telephone number, etc. may also be used as personal identification information, and this personal identification information is associated with the examinee ID and stored in the examinee information storage means 43), the teaching material ID of the purchased teaching materials, and the teaching material purchase date. Then, each time the examinee who purchased the teaching materials actually uses the teaching materials, the date and time of the teaching material use is stored in the teaching material usage status storage means 46 in association with the name, date of birth, and other personal identification information, as well as the teaching material ID. The writing of the teaching material use date and time is executed by the learning system 120 that supports learning using the teaching materials.

The teaching material use result tabulation means 36 uses the data stored in the teaching material use status storage means 46, the test information storage means 44, the examinee information storage means 43, and the results storage means 42 to tabulate the results of the use of teaching materials by each examinee who used each teaching material, and executes a process of storing the tabulation results in the teaching material use result storage means 47 in association with personal identification information such as name and date of birth, and teaching material ID. Specifically, as shown in the bottom of FIG. 8, the same type of test may be conducted multiple times, such as the first test, the second test, etc., and information is tabulated on how much each examinee's grade improved by using each teaching material while the same type of test is repeated. Note that, even if the test is the same type, different test IDs are assigned for the first test and the second test, but if the test is the same type, part of the test ID is shared so that it can be identified as the same type of test, or a table (same type test association storage means) that associates test IDs for the same type of test is prepared. In addition, the examinee ID will be an examinee number or the like, and normally, even for the same person, the examinee ID will be different for each exam, so performance data for each exam taken by the same person will be collected using personal identification information such as name and date of birth that is associated with the examinee ID and stored in the examinee information storage means 43.

That is, the educational material use result tabulation means 36 first acquires the educational material use dates and times (cumulative) shown at the bottom of FIG. 8, which are stored as cumulative data in the educational material use status storage means 46, and determines the temporal relationship with each test implementation date (test implementation date associated with the test ID for the same type of test) stored in the test information storage means 44, and identifies the "test ID for the last test (1) taken during the period before the start of educational material use" and the "test ID for the test (2) taken immediately after the last educational material use or the last test (2) while the educational material is still being used." In the tabulation example A at the bottom of FIG. 8, the fourth test corresponds to the "last test (1) taken during the period before the start of educational material use," and the sixth test corresponds to the "test (2) taken immediately after the last educational material use." In the tabulation example B, the fourth test corresponds to the "last test (1) taken during the period before the start of educational material use," and the seventh test corresponds to the "last test (2) while the educational material is still being used."

Next, the teaching material usage result tabulation means 36 uses the test ID for the tests (1) and (2) identified as above, and personal identification information such as the name and date of birth of the teaching material user to be tabulated, to obtain the examinee ID of the teaching material user corresponding to the test ID for the tests (1) and (2) from the examinee information storage means 43. The teaching material usage result tabulation means 36 then uses the test ID and the examinee ID to obtain grade data from the grade storage means 42, and stores the obtained grade data and the test ID in the teaching material usage result storage means 47 in association with the personal identification information such as the name and date of birth of the teaching material user to be tabulated, and the teaching material ID. Furthermore, the teaching material usage result tabulation means 36 calculates the difference between the performance data for tests (1) and (2) by the same person (teaching material user) stored in the teaching material usage result storage means 47, i.e., the amount of change in the performance data for test (2) relative to the performance data for test (1), and stores the calculated amount of change (assuming that the amount of change is positive) in the teaching material usage result storage means 47 as teaching material usage result data.

The advertising data creation means 37 executes a process of tallying up the material use result data of each material user stored in the material use result storage means 47 for each material, calculating various index values such as the average, median, mode, and maximum of the increased scores, creating advertising data for each material using the calculated index values, and storing the created advertising data in the advertising data storage means 48 in association with the exam ID (the exam ID for the next exam of the same type), the material ID, the material type, and the classification of the performance data of the person to whom the advertisement is to be displayed. The material type is, for example, a material for learning Chinese, a material for learning English, a law-related material for the bar exam, a material related to electrical technology, etc., and the exam ID specifies the content of the exam (for example, a Chinese language exam, a bar exam, etc.), so that the material type corresponding to the exam ID is determined. The classification of the performance data of the advertisement display target is information indicating the grade level of the test-takers to whom the advertisement is recommended for purchase, such as test-takers with poor grades, test-takers with average grades, and test-takers with relatively high grades, when the advertisement is posted after the end of the test, and is data indicating classifications such as "less than 10 points", "10 points or more and less than 30 points", "30 points or more and less than 69 points", and "70 points or more". The advertisement data created here is the advertisement data to be used in the next test (for example, the advertisement data to be posted after the end of the next test), and the advertisement data stored in the advertisement data storage means 48 is constantly updated. Specifically, for example, if the material use result data of each material user who used the material α is tallied and the average score increase is 5 points, advertisement data such as "<Material α> Supervised by Professor Y of Space University. People who used this material have increased their test scores by an average of 5 points. The usage fee is XXX yen/month" is created (see FIG. 8).

For example, after the end of the test, the advertisement posting means 38 uses the test ID for the test conducted and the test score data of the test taker in the test conducted to acquire from the advertisement data storage means 48 advertisement data for each teaching material (advertising data for teaching materials appropriate to be recommended to the test taker) associated with the classification of the performance data of the advertisement display target person corresponding to the test taker's performance data, and executes a process of posting the acquired advertisement data, for example, on the test result display screen 400 (see FIG. 8) after the end of the test. For example, as shown in FIG. 8, the result display screen 400 is provided with a result display section 410 and multiple (or one) advertisement display sections 420, and the advertisement posting means 38 displays advertisements on the advertisement display section 420 using the advertisement data acquired from the advertisement data storage means 48. Then, when a test taker who has seen the advertisement on the advertisement display section 420 clicks on the advertisement display section 420 of any of the teaching materials, the processing by the teaching material purchase acceptance means 35 is executed as already described in detail. In addition, such advertisement display may be performed not only on the display unit 63 of the non-transmissive head mounted display 60, but also in a test without using the non-transmissive HMD 60.

<The person (vendor) who supplies the test questions and the person (vendor) who monitors and properly conducts the tests are not the same: Figure 9>

In the above embodiment, the person (vendor) who supplies the test questions and the person (vendor) who monitors and properly conducts the test are the same, but these persons (vendors) do not necessarily have to be the same. For example, the test system 500 shown in FIG. 9 is operated and managed by multiple persons (vendors). This test system 500 includes a first test management server 510 operated and managed by a first administrator (e.g., a vendor that conducts various tests) and a second test management server 520 operated and managed by a second administrator (e.g., a vendor that creates test questions, sets questions, accepts answers, evaluates, etc.). The first and second test management servers 510 and 520 are each composed of one or more computers.

In such a configuration, for example, a business model can be adopted in which there are multiple (many) second administrators who supply test questions, etc., and these multiple second administrators conduct tests using questions that they have prepared, under the management and control of a single first administrator (one company) who conducts the tests, including monitoring. Note that the relationship between the first administrator and the second administrator is not limited to this one-to-many relationship, but may be many-to-many or many-to-one. In the case of many-to-one, for example, a business model can be adopted in which a single second administrator (one company) creates and provides Chinese test questions and English test questions, and tests using the respective questions are conducted under the management and control of different first administrators.

The processing means 510A of the first test management server 510 includes the pre-registration acceptance means 21, the face image acquisition means 22, the face feature extraction means 23, the wearing/removing state data receiving means 26, the face image acquisition necessity judgment means 28, the score calculation means 29, the certificate creation means 30, the light emission management means 31, the fraud judgment means 32, the monitoring means 33, and the identity verification means 34, as in FIG. 1 of the embodiment, and also includes the teaching material purchase acceptance means 35, the teaching material use result aggregation means 36, the advertisement data creation means 37, and the advertisement posting means 38, as in FIG. 8 described above. The first test management server 510 also includes the score storage means 42, the examinee information storage means 43, the test information storage means 44, and the light emission confirmation result storage means 45, as in FIG. 1 of the embodiment, and also includes the teaching material use state storage means 46, the teaching material use result storage means 47, and the advertisement data storage means 48, as in FIG. 8 described above.

The first test management server 510 further includes a wearing/removing status storage means 540, which is part of the answer storage means 41 of the above embodiment. Therefore, the answer storage means 41 of the above embodiment also serves as a wearing/removing status storage means. Specifically, as shown in FIG. 9, the wearing/removing status storage means 540 stores, in association with each other, a test ID, an examinee ID, a question ID, wearing/removing status data for a question time period (combined data of information indicating a question time period and whether the test time period is worn or not worn), and wearing/removing status data for an answer time period (combined data of information indicating an answer time period and whether the test time period is worn or not worn).

This first test management server 510 is connected via network 1 to a monitor terminal 80, an identity verification requester terminal 110, and a first administrator terminal 530 operated by a first administrator (e.g., a company that performs various tests including monitoring). The first administrator terminal 530 is a functionally divided version of the administrator terminal 100 in FIG. 1 of the embodiment.

The processing means 520A of the second test management server 520 includes a question data sending means 24, an answer data receiving means 25, and an evaluation means 27, similar to FIG. 1 of the above embodiment. The second test management server 520 also includes a question storage means 40, similar to FIG. 1 of the above embodiment. Furthermore, the second test management server 520 includes an answer storage means 541, which is part of the answer storage means 41 (see FIG. 4) of the above embodiment. Specifically, the answer storage means 541 stores a test ID, examinee ID, question ID, answer data, and evaluation result data in association with each other, as shown in FIG. 9.

The second test management server 520 is also connected via the network 1 to an evaluator terminal 90 and a second administrator terminal 531 operated by a second administrator (e.g., a company that creates test questions, sets questions, accepts answers, evaluates, etc.). The second administrator terminal 531 is a functionally divided version of the administrator terminal 100 in FIG. 1 of the embodiment.

In the case of the configuration shown in FIG. 9, the process of transmitting and receiving question data (the process of steps S5 and S6 in FIG. 4 in the above embodiment) and the process of transmitting and receiving an instruction to switch to the display mode for the action answer by the wearer's line of sight imaging device 68 (see FIG. 2 and FIG. 3 in the above embodiment) (the process of steps S9 and S10 in FIG. 4 in the above embodiment) are processed from the second test management server 520 to the examinee terminal 50. Also, the process of transmitting and receiving answer data (the process of steps S15 and S16 in FIG. 4 in the above embodiment) is processed from the examinee terminal 50 to the second test management server 520. Furthermore, the process of evaluating the answers (determining whether they are correct or not and marking them) (the process of step S17 in FIG. 4 in the above embodiment) is executed by the second test management server 520.

On the other hand, the processes of sending and receiving pre-registered data (steps S1 and S2 in FIG. 4 of the embodiment), sending and receiving person image data (steps S3 and S4 and steps S23 and S24 in FIG. 4 of the embodiment), and sending and receiving attachment/detachment state data (steps S7, S8 and steps S11 and S12 in FIG. 4 of the embodiment) are processed from the examinee terminal 50 to the first test management server 510. In addition, the processes of sending and receiving light emission instructions and light emission confirmation image data for light emission management of the light-emitting unit 67 of the non-transparent HMD 60 (see FIG. 2 and FIG. 3 of the embodiment) (steps S13 and S14 in FIG. 4 of the embodiment) are processed between the first test management server 510 and the examinee terminal 50. In addition, the processes of sending and receiving instructions to remove the non-transparent HMD 60 and the accompanying sending and receiving person image data (steps S18 to S21 in FIG. 4 of the embodiment) are also processed between the first test management server 510 and the examinee terminal 50. The score calculation process (step S25 in FIG. 4 of the embodiment), the advertisement posting process (see FIG. 8), and the certificate creation process (step S26 in FIG. 4 of the embodiment) are executed by the first test management server 510. Note that the score calculation process uses the evaluation result data stored in the answer storage means 541 of the second test management server 520, so the score calculation means 29 of the first test management server 510 acquires the evaluation result data from the second test management server 520 via the network 1.

In addition, in the case of the configuration shown in FIG. 9, the screen displayed on the display unit 63 (see FIG. 2 and FIG. 3) of the non-transparent head mounted display 60, or in the case of a test or question that does not require wearing the non-transparent HMD 60, the screen of the display means 51 (display means excluding the non-transparent HMD 60) of the examinee terminal 50 will display the administrative screen transmitted from the first test management server 510 and the screen for presenting questions transmitted from the second test management server 520. In this case, the administrative screen transmitted from the first test management server 510 may be the outer frame portion, and the screen for presenting questions transmitted from the second test management server 520 may be displayed within the frame, or the administrative screen and the screen for presenting questions may be switched and displayed. In the latter case, for example, the screen immediately before the switch may include information (URL, etc.) of the screen immediately after the switch. The screen display process and associated processes on the examinee terminal 50 may be executed, for example, by a web browser and a program (e.g., JavaScript (ECMAScript), etc.: JavaScript is a registered trademark) provided by the first test management server 510 or the second test management server 520 in association with the web page, or a dedicated program with browsing functionality may be installed and executed on the examinee terminal 50 (main body 50A of the examinee terminal 50 or main body 60A of the non-transparent HMD 60).

<Configuration for issuing electronic certificates and recording them in a distributed ledger>

In addition, in the above embodiment, the certificate creation means 30 is configured to issue a certificate 300 that includes a facial photograph 310 and two-dimensional barcodes 321-325 (however, whether or not to include them can be selected) as shown in FIG. 7, but an electronic certificate may be issued at the request of the examinee, and stored in a distributed ledger (e.g., a blockchain) in association with the examinee ID or an alternative identification information for storage, making it difficult to tamper with the data. When the examinee needs it (when the examinee requests it), the examinee's electronic certificate may be obtained from the distributed ledger using the identification information such as the examinee ID and sent to the examinee's terminal 50 (however, it may be a terminal different from the one used at the time of the examination). In this case, the electronic certificate may be registered in the distributed ledger and obtained from the distributed ledger by accessing a constituent node (one of multiple computers connected by a peer-to-peer network for forming the distributed ledger) of the distributed ledger (e.g., a blockchain) from the examination management server 20, or the examination management server 20 itself may be a constituent node.

<A configuration that uses a distributed ledger to handle virtual currencies for test takers>

As described above, when the constituent nodes of the distributed ledger (such as a block chain) are accessible from the test management server 20 or the test management server 20 itself is a constituent node, the test management server 20 may be configured to handle virtual currency for examinees. Specifically, when the performance data calculated by the performance calculation means 29 and stored in the performance storage means 42 is good (when it is equal to or greater than a predetermined threshold) or when the test system 10 is used for a long time (when it is equal to or greater than a predetermined threshold), a service may be provided in which virtual currency is granted to the examinee. In this case, a threshold may be set in stages, and the amount of virtual currency granted may be increased or decreased according to the performance or the length of use time. In other words, the better the performance or the longer the use time, the more virtual currency may be granted. In addition, when the performance is good or the use time is long, instead of granting virtual currency, points that can be used to purchase teaching materials or points that can be used to purchase or rent the non-transparent head-mounted display 60 may be granted to the examinee. More specifically, an examinee assistance means (not shown) provided in the exam management server 20 performs calculations for awarding virtual currency and points for each examinee, the awarded virtual currency is recorded in the examinee's account (e.g., an Ethereum address, etc.) in a distributed ledger (such as a blockchain), and the awarded points are stored in a point storage means (not shown) provided in the exam management server 20. Note that if the examinee's bank account information or credit card information is stored in the examinee information storage means 43, real currency may be transferred instead of awarding virtual currency. This transfer process is also performed by the examinee assistance means (not shown).

<The system uses data on the state of the device to estimate usage time and awards virtual currency, points, and real currency.>

Furthermore, as described above, when virtual currency, points, or real currency is awarded according to the usage time, the use time may be estimated using the wearing/removing state data stored in the answer storage means 41. Specifically, the examinee assistance means (not shown) may estimate the time length based on the number of pieces of wearing/removing state data acquired, assuming that the wearing/removing state data is acquired at regular time intervals, or may award virtual currency, points, or real currency according to the size of the number of pieces of wearing/removing state data acquired itself. In this case, the fraud determination means 32 executes a process to determine whether the wearing/removing state is appropriate for the question type of the question, using the wearing/removing state data for the question time period and the wearing/removing state data for the answer time period received by the wearing/removing state data receiving means 26 and stored in the answer storage means 41, so the examinee assistance means (not shown) may count only the cases in which the wearing/removing state is judged to be appropriate by the fraud determination means 32, and use this as the number of wearing/removing state data (the number itself, or the length of time estimated from the number). In other words, attachment/detachment status data acquired under circumstances where there is a possibility of fraud may not be counted as usage.

<Configuration that uses human image data to detect unnatural movements of examinees while processing questions that do not require wearing a non-transparent HMD 60>

In the above embodiment, during the processing of a question that requires the wearing of the non-transparent head mounted display 60, the cheating determination means 32 judges whether or not cheating is possible using the wearing/removing state data received by the wearing/removing state data receiving means 26 and stored in the answer storage means 41, and transmits information that cheating is possible (for example, information that the monitoring level requires a warning display using a red or yellow frame, etc.) to the monitoring means 33. However, during the processing of a question that does not require the wearing of the non-transparent head mounted display 60, for example, the following processing can be performed. That is, the monitoring means 33 in the above embodiment receives person image data including the face of the examinee that is captured by the person imaging device 53 provided in the examinee terminal 50 and transmitted via the network 1, and executes a process of displaying the received person image data as a video on the monitoring screen 200 of the display means 81 of the monitor terminal 80 as shown in FIG. 6. At this time, the monitoring means 33 may pass the received person image data to the cheating determination means 32. The cheating determination means 32 may use the received person image data to detect unnatural movements of the examinee (for example, movements that may indicate cheating, such as frequently turning the head to the side or frequently looking to the side), and if detected, may transmit information to the monitoring means 33 that there is a possibility of cheating (for example, information that the monitoring level requires a warning display using a red or yellow frame, etc.).

<Configuration for calculating identity using facial feature score, voiceprint score, etc.>

In the above embodiment, the fraud determination means 32 determines whether all of the multiple facial features extracted by the facial feature extraction means 23 and stored in the examinee information storage means 43 match or approximately match, and if they do not match or approximately match, executes a process to determine that fraud has occurred. Furthermore, if the fraud determination means 32 determines that all of the multiple facial features match or approximately match, the certificate creation means 30 executes a process to adopt the facial image data among the multiple facial image data in which the examinee's face is most clearly shown as the facial photograph to be published on the certificate 300 (see Figure 7). However, the fraud determination means and certificate creation means of the present invention may also be configured to calculate the degree of identity using the facial feature score, voiceprint score, etc., as follows, and determine whether or not to issue the certificate 300 depending on the degree of identity.

In other words, in such a configuration (hereinafter referred to as "a configuration for calculating identity using facial feature score, voiceprint score, etc."), the facial image acquisition means executes a process of acquiring facial image data from person image data including the examinee's face that is captured by the person imaging device 53 provided on the examinee terminal 50 and transmitted via the network 1 at the start of the test, during the period during the test when the wearing/removal status data indicates non-wearing, or at the start of the test, during the period during the test, and at the end of the test.

More specifically, at the start of the test, the face image acquisition means stores the face image data acquired from the person image data in the examinee information storage means 43 in association with the examinee ID and the test ID as face image data for the face photo to be published on the certificate 300 (see FIG. 7) created by the certificate creation means. Therefore, in the above embodiment, the face image data in which the examinee's face is most clearly shown among the multiple face image data acquired by the face image acquisition means 22, that is, one of the multiple face image data, was used as the face photo 310 to be published on the certificate 300 (see FIG. 7). However, when the face feature score, voiceprint score, etc. are used to calculate the identity, the face image data acquired at the start of the test (including the stage before the start of the test, as already described in detail. In the example of FIG. 4, at the time of steps S3 and S4) will necessarily be the face image data for the face photo 310 to be published on the certificate 300, so it is essential to acquire face image data at the start of the test (including before the start of the test).

Furthermore, during the time period during the test or at the end of the test (including after the test, as already described in detail), the facial image acquisition means associates the facial image data acquired from the person image data with the examinee ID and the test ID as facial image data for fraud determination in the examinee information storage means 43 only when the facial feature amount score calculated by the facial feature amount extraction means described below is less than or equal to a predetermined threshold amount (when the similarity of the facial feature amount is low). On the other hand, when the facial feature amount score is equal to or greater than the predetermined threshold amount (when the similarity of the facial feature amount is high), the facial image data acquired from the person image data is not stored in the examinee information storage means 43.

If there are many cases where the facial feature score is low (when the similarity of the facial features is low) as in the former case, the facial feature average score described below will be low, and therefore the degree of identity described below will be low, which may result in a situation where the certificate 300 (see Figure 7) is not issued. If such a situation occurs and the examinee claims that they have not committed any fraud, the facial image data stored in the examinee information storage means 43 will be used to manually determine whether or not there has been any fraud. On the other hand, if there are many cases where the facial feature score is high (when the similarity of the facial features is high) as in the latter case, the facial image data is not stored in the examinee information storage means 43, thereby reducing the amount of data to be stored.

The facial feature extraction means executes a process of extracting facial features for each of the multiple facial image data acquired by the facial image acquisition means described above.

Specifically, at the start of the test (including before the test starts), the facial feature extraction means extracts facial features from the facial image data acquired by the facial image acquisition means described above, and stores the extracted facial features in the examinee information storage means 43 in association with the examinee ID and the test ID as reference data for calculating the identity degree. Also, during the time period during the test or at the end of the test (including after the test is finished), the facial feature extraction means extracts facial features from the facial image data acquired by the facial image acquisition means described above, and stores the extracted facial features in the examinee information storage means 43 in association with the examinee ID and the test ID as matching data for calculating the identity degree, and executes a process of calculating a facial feature score indicating the similarity between the facial features as matching data for calculating the identity degree and the facial features as reference data for calculating the identity degree stored in the examinee information storage means 43. The facial feature score calculated here may be stored in the examinee information storage means 43 in association with the examinee ID and test ID along with the facial feature as matching data for calculating the identity, or it may not be stored on the assumption that it will be calculated again by the fraud determination means.

The pre-registration reception means is the same as the pre-registration reception means 21 of the above embodiment, and is therefore given the same reference numerals. The pre-registration reception means 21 receives voice data for voiceprint extraction that is acquired by the microphone 55 of the examinee terminal 50 or the microphone 70 of the non-transparent head-mounted display 60 and transmitted via the network 1 at the start of the test (including before the test starts. In the example of FIG. 4, at the time of steps S1 and S2), performs voiceprint analysis using the received voice data, and executes a process of storing the obtained voiceprint data in the examinee information storage means 43 in association with the examinee ID and the test ID as reference data for calculating the degree of identity. In this case, the voiceprint analysis method can be an existing method, and may be a text-dependent method that uses voice data obtained by vocalizing a fixed phrase, or a text-independent method that uses voice data of a non-fixed utterance.

The pre-registration reception means 21 also receives iris data captured by the person imaging device 53 or auxiliary imaging device 54 provided on the examinee terminal 50, or the inward imaging device 69 of the non-transmissive HMD 60, at the start of the test (including before the test starts; in the example of FIG. 4, at the time of steps S1 and S2) and transmitted via the network 1, and executes a process of storing the received iris data in the examinee information storage means 43 in association with the examinee ID and test ID as reference data for calculating the degree of identity.

Furthermore, the pre-registration reception means 21 receives fingerprint data that is read by the fingerprint reader 57 provided on the examinee terminal 50 and transmitted via the network 1 at the start of the test (including before the test starts; in the example of FIG. 4, at the time of steps S1 and S2), and executes a process of storing the received fingerprint data in the examinee information storage means 43 in association with the examinee ID and test ID as reference data for calculating the degree of identity.

In addition, at the start of the exam (including before the exam starts; in the example of Figure 4, at the time of steps S1 and S2), the pre-registration reception means 21 responds to a request from the examinee terminal 50 by transmitting display data for a pre-registration screen (e.g., a web screen) via the network 1, and receives personal information such as the examinee's address, name, date of birth, hobbies, occupation, student ID number, etc. transmitted from the examinee terminal 50 via the network 1, and executes a process of storing the received personal information in the examinee information storage means 43 in association with the automatically assigned examinee ID (examination number, etc.) and exam ID as reference data for calculating the degree of identity.

In the case where the identity degree is calculated using the facial feature score, the voiceprint score, etc., unlike the above embodiment, the test management server 20 is provided with a matching data acquisition means. This matching data acquisition means receives voice data for voiceprint extraction that is acquired by the microphone 55 of the examinee terminal 50 or the microphone 70 of the non-transparent head mounted display 60 during the test period or at the end of the test (including the end of the test) in addition to the test period, and is transmitted via the network 1, performs voiceprint analysis using the received voice data, and executes a process of storing the obtained voiceprint data in the examinee information storage means 43 in association with the examinee ID and the test ID as matching data for identity degree calculation. In this case, the voiceprint analysis method can be an existing method, and may be a text-dependent method using voice data obtained by vocalizing a fixed phrase, or a text-independent method using voice data of a non-fixed phrase. When answer data inputted aloud is used, the text-independent method is used.

The matching data acquisition means also receives iris data captured by the person imaging device 53 or auxiliary imaging device 54 provided on the examinee terminal 50, or the inward imaging device 69 of the non-transmissive HMD 60, during the time period during the test or at the end of the test (including after the test is completed) and transmitted via the network 1, and executes a process of storing the received iris data in the examinee information storage means 43 in association with the examinee ID and test ID as matching data for calculating identity.

The matching data acquisition means further receives fingerprint data that is read by the fingerprint reader 57 provided on the examinee terminal 50 and transmitted via the network 1 during the test period or at the end of the test (including after the test period) in addition to during the test period, and executes a process of storing the received fingerprint data in the examinee information storage means 43 in association with the examinee ID and test ID as matching data for calculating the degree of identity.

The matching data acquisition means also transmits display data for a matching screen (e.g., a web screen) to the examinee terminal 50 via the network 1 during the time period during the test or at the end of the test (including after the test has ended) in addition to during the time period during the test, and receives personal information such as the examinee's address, name, date of birth, hobbies, occupation, student ID number, etc. transmitted from the examinee terminal 50 via the network 1, and executes a process of storing the received personal information in the examinee information storage means 43 in association with the examinee ID and test ID as matching data for calculating the degree of identity.

The fraud determination means calculates multiple facial feature scores, multiple voiceprint scores, multiple iris scores, multiple fingerprint scores, and multiple personal information scores (e.g., multiple hobby scores, etc.), averages them to calculate an average facial feature score, an average voiceprint score, an average iris score, an average fingerprint score, and an average personal information score (e.g., an average hobby score, etc.), and further integrates these average scores (simple average, weighted average, harmonic mean, etc.) to calculate a degree of identity (a number indicating the likelihood of identity), and executes a process to determine that fraud has occurred if the calculated degree of identity is low (below or equal to a predetermined threshold value).

Specifically, the fraud determination means uses each of the facial features stored in the examinee information storage means 43 as matching data for calculating the identity degree and the facial features as reference data for calculating the identity degree to calculate multiple facial feature scores indicating the similarity of those facial features (the degree to which each matching data is similar to the reference data). Note that if the facial feature scores are calculated by the above-mentioned facial feature extraction means and stored in the examinee information storage means 43 together with the facial features (facial features as matching data for calculating the identity degree) in association with the examinee ID and the test ID, multiple facial feature scores may be obtained from the examinee information storage means 43. The facial feature scores, i.e., the numerical values indicating the similarity of the facial features, may be numerical values used in existing face authentication, such as likelihood or probability indicating the certainty that the person is the actual person. Furthermore, the fraud determination means calculates a facial feature average score by averaging (a simple average may be used) the multiple calculated or acquired facial feature scores.

The fraud determination means also uses each of the voiceprint data stored in the examinee information storage means 43 as matching data for calculating the identity degree, and the voiceprint data as reference data for calculating the identity degree, to calculate a number of voiceprint scores indicating the similarity of the voiceprint data (the degree to which each of the matching data resembles the reference data). The voiceprint score, i.e., the numerical value indicating the similarity of the voiceprint data, may be a numerical value used in existing voiceprint authentication, such as a likelihood or probability indicating the certainty that the person is the real person. Furthermore, the fraud determination means calculates an average voiceprint score by averaging (a simple average may be used) the multiple calculated voiceprint scores.

Furthermore, the fraud determination means calculates multiple iris scores indicating the similarity of the iris data (the degree to which each piece of matching data resembles the reference data) using each piece of iris data stored in the examinee information storage means 43 as matching data for calculating the identity of the examinee, and iris data as reference data for calculating the identity of the examinee. The iris score, i.e., the numerical value indicating the similarity of the iris data, may be a numerical value used in existing iris authentication, such as a likelihood or probability indicating the certainty that the examinee is the actual examinee. Furthermore, the fraud determination means calculates an average iris score by averaging (a simple average may be used) the multiple calculated iris scores.

The fraud determination means also uses each of the fingerprint data stored in the examinee information storage means 43 as matching data for calculating identity, and the fingerprint data as reference data for calculating identity, to calculate a number of fingerprint scores indicating the similarity of the fingerprint data (the degree to which each of the matching data resembles the reference data). The fingerprint score, i.e., the numerical value indicating the similarity of the fingerprint data, may be a numerical value used in existing fingerprint authentication, such as a likelihood or probability indicating the certainty that the person is the real person. Furthermore, the fraud determination means calculates an average fingerprint score by averaging (a simple average may be used) the multiple calculated fingerprint scores.

Furthermore, the fraud determination means calculates a number of personal information scores (e.g., a number of hobby scores, etc.) indicating the similarity of the personal information (the degree to which each piece of matching data resembles the reference data) using each piece of personal information (e.g., hobbies, etc.) stored in the examinee information storage means 43 as matching data for calculating the identity degree and personal information (e.g., hobbies, etc.) as reference data for calculating the identity degree. For example, when hobbies are input using options, it is whether or not the selected options match, and when hobbies are input using text, playing the piano and playing musical instruments are relatively similar, so the hobby score will be relatively high. Furthermore, the fraud determination means calculates a personal information average score (e.g., a hobby average score, etc.) by averaging (a simple average may be used) the calculated number of personal information scores (e.g., a number of hobby scores, etc.).

The fraud determination means then calculates the identity degree by integrating the facial feature average score, voiceprint average score, iris average score, fingerprint average score, and personal information average score (e.g., hobby average score, etc.), and stores the calculated identity degree in the examinee information storage means 43 in association with the examinee ID and test ID. At this time, the integration can be performed, for example, by a simple average, weighted average, harmonic average, etc. The fraud determination means also compares the calculated identity degree with a predetermined threshold, and executes a process of determining that fraud has occurred if the calculated identity degree is low (below or equal to the threshold). Note that average scores other than the "facial feature average score and voiceprint average score" are not essential and do not have to be used for integration.

The certificate creation means executes a process to create an electronic or paper certificate 300 (see FIG. 7) to be issued to the examinee. In this respect, it is similar to the certificate creation means 30 in the above embodiment.

On the other hand, the certificate creation means 30 of the above embodiment is configured to execute a process of adopting the facial image data in which the examinee's face is most clearly shown among the facial image data acquired by the facial image acquisition means 22 as the facial photograph 310 to be published on the certificate 300 (see FIG. 7) when the fraud determination means 32 determines that all of the facial features match or approximately match. In contrast, when the identity degree is calculated using the facial feature amount score, voiceprint score, etc., the certificate creation means does not create the certificate 300 (see FIG. 7) when the identity degree calculated by the fraud determination means is equal to or less than a predetermined threshold (when it is determined that fraud has occurred). Also, when the identity degree calculated by the fraud determination means exceeds or is equal to or greater than a predetermined threshold, the certificate 300 (see FIG. 7) is created, and at this time, the facial image data acquired by the facial image acquisition means at the start of the test (including before the start of the test) is used as the facial image data for the facial photograph 310 to be published on the certificate 300.

<Configuration for issuing a voice command in addition to an instruction to remove the transparent head-mounted display 60>

In the above embodiment, the test management server 20 is provided with a facial image acquisition necessity determination means 28, but if the configuration is such that identity is calculated using a facial feature score, a voiceprint score, etc., it is preferable to also provide a voiceprint acquisition necessity determination means. This voiceprint acquisition necessity determination means executes a process to send a voice instruction to the examinee terminal 50 via the network 1 at any timing between questions in order to obtain voice data for voiceprint extraction during the time period during the test when questions other than those to be answered aloud have appeared in succession for a predetermined number of questions or for a predetermined period of time.

The facial image acquisition means 22 in the above embodiment is configured to execute a process of acquiring facial image data from the person image data transmitted from the examinee terminal 50 via the network 1 when the facial image acquisition necessity determination means 28 sends an instruction to the examinee terminal 50 to remove the non-transparent head mounted display 60, and when it is determined that the attachment/detachment status data after the removal instruction indicates that the display is not being worn. Based on a similar concept, it is preferable that the matching data acquisition means be configured as follows.

In other words, when the voiceprint acquisition necessity determination means sends a vocal instruction to the examinee terminal 50, the matching data acquisition means preferably executes a process to receive voice data for voiceprint extraction sent from the examinee terminal 50 via the network 1, perform a voiceprint analysis using the received voice data, and store the obtained voiceprint data in the examinee information storage means 43 in association with the examinee ID and test ID as matching data for calculating the identity degree.

As described above, the testing system and program of the present invention are suitable for use in conducting various types of tests, such as qualification tests, certification tests, qualification tests, entrance exams, employment exams, graduation exams, completion exams, final exams, promotion exams, mock exams, and daily learning effectiveness check tests, to measure abilities and knowledge such as language ability, conversation ability, communication ability, and skills.

LIST OF SYMBOLS 1 Network 10,500 Examination system 20,510,520 Examination management server 22 Face image acquisition means 23 Face feature extraction means 24 Question data transmission means 25 Answer data reception means 26 Wearing/removing state data reception means 27 Evaluation means 28 Face image acquisition necessity determination means 30 Certificate creation means 31 Light emission management means 32 Fraud determination means 33 Monitoring means 50 Examinee terminal 60 Non-transmissive head mounted display 63 Display unit 66 Wearing/removing detection device 67 Light emission unit 68 Wearer's line of sight imaging device 80 Monitor terminal

Claims (13)

A testing system configured with a computer that administers tests to examinees in remote locations via a network,
A test management server is provided which is connected to an examinee terminal operated by an examinee via a network,
This test management server is
a question data transmission means for executing a process of transmitting test question data to the examinee terminal via a network in order to display the test question data on a display unit of a non-transmissive head mounted display worn by the examinee;
an answer data receiving means for receiving the answer data of the examinees transmitted from the examinee terminals via a network;
a wearing/removing state data receiving means for receiving wearing/removing state data of the non-transmissive head mounted display, the wearing/removing state data being detected by a wearing/removing detection device provided in the non-transmissive head mounted display and being transmitted from the examinee terminal via a network;
and a fraud determination means for executing a process to determine whether the attachment/detachment state is appropriate for the question type of the question, using the attachment/detachment state data for the question time period and the attachment/detachment state data for the answer time period received by the attachment/detachment state data receiving means. an observer terminal connected to the test management server and displaying information for an observer to observe the test;
The test management server includes:
a monitoring means for receiving image data including the face of the examinee captured by a person imaging device provided in the examinee terminal and transmitted via a network, and for executing a process of displaying the received image data as a moving image on the monitor terminal;
The fraud determination means includes:
during processing of a question that requires the wearing of the non-transparent head mounted display, if it is determined that the wearing/removal state data during the question time period indicates that the non-transparent head mounted display is worn and the wearing/removal state data during the answer time period indicates that the non-transparent head mounted display is not worn, a process is executed to transmit information that there is a possibility of cheating to the monitoring means,
The monitoring means includes:
The test system of claim 1, further comprising a configuration for receiving a signal from the fraud determination means and executing a warning display process around or in a peripheral area of the examinee's portrait image data displayed on the monitor terminal. The test management server includes:
a face image acquisition means for acquiring face image data from person image data including the face of the examinee captured by a person image capture device provided in the examinee terminal and transmitted via a network at the start and/or end of the test, and during the time period during the test when the wearing/removal status data indicates non-wearing;
a facial feature extraction unit that executes a process of extracting facial features from each of the plurality of facial image data acquired by the facial image acquisition unit;
a certificate generating means for executing a process of generating an electronic or paper certificate to be issued to an examinee;
The fraud determination means includes:
The method is configured to determine whether or not all of the facial feature amounts extracted by the facial feature amount extraction means match or substantially match, and if all of the facial feature amounts do not match or substantially match, to determine that fraud has occurred;
The certificate generating means includes:
The examination system described in claim 1 or 2, characterized in that if the fraud determination means determines that all of the multiple facial features match or approximately match, a process is executed to adopt the facial image data among the multiple facial image data which most clearly shows the examinee's face as the facial photograph to be printed on the certificate. The test management server includes:
a facial image acquisition necessity determination means for executing a process of transmitting an instruction to remove the non-transmissive head mounted display to the examinee terminal via a network at any timing in order to obtain facial image data for the time period during the test, when questions requiring the wearing of the non-transmissive head mounted display occur consecutively for a predetermined number of questions or for a predetermined period of time, and when it is determined that the wearing/removal state data during the consecutive questions indicates that the non-transmissive head mounted display is being worn;
The face image acquisition means includes:
The examination system according to claim 3, characterized in that when the facial image acquisition necessity determination means sends an instruction to the examinee terminal to remove the non-transparent head mounted display and when it is determined that the attachment/detachment status data after the removal instruction indicates that the display is not being worn, the examination system is configured to execute a process of acquiring facial image data from the person image data sent from the examinee terminal via a network. The non-transmissive head mounted display is provided with a wearer's line of sight imaging device capable of imaging a space outside the non-transmissive head mounted display with a line of sight of a wearer of the non-transmissive head mounted display,
The problem data transmission means includes:
during processing of a question that requires wearing the non-transmissive head mounted display and that requires an answer involving a movement, at a timing when a question period transitions from a question period to an answer period, a process is also executed to transmit, via a network to the examinee terminal, an instruction to switch to a display mode for answers involving a movement, in which wearer's gaze image data captured by the wearer's gaze imaging device is displayed as a moving image on at least a part of a display unit of the non-transmissive head mounted display;
The answer data receiving means includes:
The test subject terminal is configured to execute a process of receiving the wearer's gaze image data transmitted from the test subject terminal via a network as the answer data,
The monitoring means includes:
The test system according to claim 2, further comprising a configuration for executing a process of displaying the wearer's gaze image data transmitted from the examinee terminal via a network as a moving image on the monitor terminal, either in switching over to or together with the display of the person image data. The test management server includes:
The test system according to claim 5, further comprising an evaluation means for performing character reading processing using the wearer's gaze image data received as the answer data by the answer data receiving means for questions to be answered in writing. The test management server includes:
The test method includes: a test device for testing a test item to be answered by performing an action in a real space using an actual test tool; and an evaluation device for performing a pattern recognition process using the wearer's gaze image data received as the answer data by the answer data receiving device.
This evaluation tool is
6. The test system according to claim 5, wherein the pattern recognition process is executed using a model for pattern recognition obtained by a learning process using wearer gaze image data determined to be correct and wearer gaze image data determined to be incorrect, as well as the labels of correct and incorrect answers corresponding thereto. The non-transmissive head mounted display is provided with a wearer's line of sight imaging device capable of imaging a space outside the non-transmissive head mounted display with a line of sight of a wearer of the non-transmissive head mounted display,
The problem data transmission means includes:
during processing of a question that requires wearing the non-transmissive head mounted display and for which written notes are permitted during the question setting time period and the answering time period, test question data is displayed on the display unit of the non-transmissive head mounted display, and a process is also executed to transmit to the examinee terminal via a network an instruction to switch to a memo display mode that displays wearer's gaze image data captured by the wearer's gaze imaging device so that the examinee can take notes in writing;
The monitoring means includes:
The test system according to claim 2, further comprising a configuration for executing a process of displaying the wearer's gaze image data transmitted from the examinee terminal via a network as a moving image on the monitor terminal, either in switching over to or together with the display of the person image data. The non-transmissive head mounted display is provided with a light emitting unit,
The test management server includes:
a light emission management means for transmitting a light emission instruction to the examinee terminal via a network for illuminating the light emitting unit at any timing during the processing of a question requiring the wearing of the non-transmissive head mounted display, and for receiving, at the timing when the light emission is performed, human image data which is captured by a human image capturing device provided in the examinee terminal and transmitted via the network as light emission confirmation image data, and for executing a process of determining whether or not light emission from the light emitting unit is captured in the received light emission confirmation image data;
The fraud determination means includes:
The test system according to any one of claims 1 to 8, characterized in that, when the light emission management means is unable to confirm the light emission in the light emission confirmation image data, a process is executed to determine that fraud has occurred. an observer terminal connected to the test management server and displaying information for an observer to observe the test;
The test management server includes:
a monitoring means for receiving image data including the face of the examinee captured by a person imaging device provided in the examinee terminal and transmitted via a network, and for executing a process of displaying the received image data as a moving image on the monitor terminal;
This monitoring means:
The test system according to claim 9, characterized in that a plurality of human image display units provided on the screen of the monitor terminal display moving image data of each of a plurality of examinees, and a plurality of light emission time zone display units provided on the screen of the monitor terminal corresponding to each of the plurality of human image display units display a light emission time zone indicating that it is a light emission time zone in synchronization with the light emission of the light emission unit of the non-transmissive head mounted display when a notification that the light emission instruction has been sent from the light emission management means is received. The test management server includes:
a face image acquisition means for acquiring face image data from person image data including the face of the examinee captured by a person image capture device provided in the examinee terminal and transmitted via a network at the start of the test, during a period during the test when the wearing/detaching state data indicates non-wearing, or at the start of the test, during the period during the test, and at the end of the test;
a facial feature extraction unit that executes a process of extracting facial features from each of the plurality of facial image data acquired by the facial image acquisition unit;
A certificate generating means for executing a process of generating an electronic or paper certificate to be issued to an examinee;
a pre-registration receiving means for receiving, at the start of a test, voice data for voiceprint extraction that is acquired by a microphone provided on the test taker terminal or a microphone on the non-transparent head mounted display and transmitted over a network, performing a voiceprint analysis using the received voice data, and storing the obtained voiceprint data in a test taker information storage means in association with a test taker ID and a test ID as reference data for calculating identity;
a verification data acquisition means for receiving voice data for voiceprint extraction, which is acquired by a microphone on said examinee terminal or a microphone on said non-transparent head mounted display and transmitted via a network during a period during the test or at the end of the test in addition to during the period during the test, performing a voiceprint analysis using the received voice data, and storing the obtained voiceprint data in said examinee information storage means in association with an examinee ID and a test ID as verification data for calculating identity;
The facial feature extraction means
At the start of the test, the facial feature amount extracted from the facial image data acquired by the facial image acquisition means is stored in the examinee information storage means as reference data for calculating the degree of identity in association with the examinee ID and the test ID;
during the test period or at the end of the test in addition to during the test period, a facial feature amount extracted from the facial image data acquired by the facial image acquisition means is stored in the examinee information storage means as matching data for calculating the identity degree in association with an examinee ID and an exam ID, and a facial feature amount score indicating a similarity between the facial feature amount as matching data for calculating the identity degree and the facial feature amount as reference data for calculating the identity degree stored in the examinee information storage means is executed;
The face image acquisition means includes:
At the start of the examination, face image data acquired from the person image data is stored in the examinee information storage means as face image data for a face photo to be printed on the certificate created by the certificate creation means, in association with the examinee ID and the examination ID;
during the test period or at the end of the test in addition to the test period, only if the facial feature amount score calculated by the facial feature amount extraction means is less than or equal to a predetermined threshold value, facial image data obtained from the person image data is associated with an examinee ID and an exam ID and stored in the examinee information storage means as facial image data for determining fraud, and if the facial feature amount score is equal to or greater than a predetermined threshold value, a process is executed not to store the facial image data obtained from the person image data in the examinee information storage means;
The fraud determination means includes:
calculate a plurality of facial feature amount scores indicating the degree of similarity between each of the facial feature amounts serving as the matching data for calculating the identity degree stored in the examinee information storage means and the facial feature amount serving as the reference data for calculating the identity degree, or obtain a plurality of the facial feature amount scores calculated by the facial feature amount extraction means and stored in the examinee information storage means in association with the examinee ID and the test ID, and calculate an average facial feature amount score by averaging the plurality of the calculated or obtained facial feature amount scores;
Furthermore, a plurality of voiceprint scores are calculated which indicate the degree of similarity between each of the voiceprint data as the verification data for calculating the identity degree stored in the examinee information storage means and the voiceprint data as the reference data for calculating the identity degree, and a voiceprint average score is calculated by averaging the calculated plurality of voiceprint scores,
a process of calculating a degree of identity by integrating the calculated facial feature amount average score and the calculated voiceprint average score,
The certificate generating means includes:
The test system according to claim 1 or 2, characterized in that the certificate is not created if the identity calculated by the fraud determination means is equal to or less than a predetermined threshold value. The test management server includes:
a facial image acquisition necessity determination means for executing a process of transmitting an instruction to remove the non-transmissive head mounted display to the examinee terminal via a network at any timing in order to obtain facial image data for the time period during the test, when questions requiring the wearing of the non-transmissive head mounted display occur consecutively for a predetermined number of questions or for a predetermined period of time, and when it is determined that the wearing/removal state data during the consecutive questions indicates that the non-transmissive head mounted display is being worn;
a voiceprint acquisition necessity determining means for executing a process of transmitting a voice instruction to said examinee terminal via a network at any timing between questions in order to obtain voice data for voiceprint extraction during the time period during the test, when questions other than those to be answered aloud have appeared in succession for a predetermined number of questions or for a predetermined period of time,
The face image acquisition means includes:
when the face image acquisition necessity determination means transmits an instruction to the examinee terminal to remove the non-transmissive head mounted display, and when it is determined that the wearing/removal state data after the removal instruction indicates that the display is not being worn, a process of acquiring face image data from the person image data transmitted from the examinee terminal via a network is executed;
The verification data acquisition means includes:
The examination system according to claim 11, characterized in that, when the voiceprint acquisition necessity determination means transmits the spoken instruction to the examinee terminal, a voiceprint analysis is performed using voice data for voiceprint extraction transmitted from the examinee terminal via a network, and a process of obtaining voiceprint data is executed. A program for causing a computer to function as a test management server constituting a test system according to any one of claims 1 to 12.

Images