Authentication device, control method of authentication device, and program
The authentication device addresses the challenge of accurately determining spoofing by presenting a stimulus to induce unconscious physical reactions and detecting these reactions, thereby enhancing the accuracy and robustness of spoofing detection in person authentication.
Patent Information
- Application Number
- JP2023196784
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-11-20
- Publication Date
- 2025-05-30
AI Technical Summary
Existing person authentication methods, such as those described in Patent Document 1 and Patent Document 2, face challenges in accurately determining spoofing, as they may misjudge genuine users with wandering attention or be fooled by repeated video presentations with multiple eye movements.
An authentication device that presents a predetermined stimulus inducing an unconscious physical reaction, detects this reaction, and determines whether the user is performing a spoofing act based on the detected reaction, thereby enhancing the accuracy of spoofing detection.
The proposed solution enables accurate determination of spoofing, reducing the risk of false positives and negatives, and ensures robustness against sophisticated spoofing attempts without burdening the user with prolonged attention requirements.
Smart Images

Figure 2025083100000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a person authentication process.
Background Art
[0002] In recent years, face recognition technology has been widely used from passing through gates to verifying the identity for electronic payments. On the other hand, there has been an act of "impersonation" such as presenting a face image of another person using a smartphone, which has become an issue. In order to prevent the act of impersonation, a technology for determining impersonation has been proposed.
[0003] Patent Document 1 displays an image that attracts the user's attention on a display, estimates the line of sight from the face of the user captured by a camera, and determines impersonation based on whether the estimated line of sight is looking at the image.
[0004] Patent Document 2 determines impersonation based on the temporal change of the user's line of sight when the display position of a series of messages displayed on the display is changed during the authentication process.
Prior Art Documents
Patent Documents
[0005]
Patent Document 1
Patent Document 2
Summary of the Invention
Problems to be Solved by the Invention
[0006] However, there was a risk that spoofing could not be accurately determined. For example, in the method described in Patent Document 1, there was a possibility that a video including a plurality of eye movements might be repeatedly presented until it matched the expected eye movement, resulting in a determination that it was not spoofing. Also, in the method described in Patent Document 2, the user needed to continuously gaze at an image or message on the display consciously, and if the user's attention wandered during the authentication procedure, there was a possibility that it might be misjudged as spoofing.
[0007] The present invention has been made in view of the above problems, and an object thereof is to enable accurate determination of spoofing.
Means for Solving the Problems
[0008] In order to achieve the above object, an authentication device as one aspect of the present invention is an authentication device for authenticating a person, and includes a presentation means for presenting a predetermined stimulus that induces an unconscious physical reaction of the person to be authenticated to the person to be authenticated, a detection means for detecting the physical reaction of the person to be authenticated after presenting the stimulus, a determination means for determining whether the person to be authenticated is performing a spoofing act based on the physical reaction detected by the detection means, and a determination means for determining whether the person to be authenticated can be authenticated based on the determination result of the determination means.
Effects of the Invention
[0009] According to the present invention, spoofing can be accurately determined.
Brief Description of the Drawings
[0010]
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Mode for Carrying Out the Invention
[0011] Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. Note that the following embodiments do not limit the invention according to the claims. Although a plurality of features are described in the embodiments, not all of these plurality of features are essential for the invention, and the plurality of features may be arbitrarily combined.
[0012] Also, in the following embodiments, it will be described as performing person authentication using a person's face, but the present invention is applicable even when performing authentication using a person's upper body or whole body. Further, in the accompanying drawings, the same or similar configurations are given the same reference numerals, and redundant descriptions are omitted.
[0013] <Embodiment 1> FIG. 1 is a block diagram showing the hardware configuration of the authentication device 100 in the present embodiment. The authentication device 100 includes a CPU 101, a ROM 102, a RAM 103, a secondary storage device 104, an imaging device 105, an input device 106, a network I / F 108, a lighting device 110, and a sound output device 112.
[0014] The CPU 101 is a central processing unit that controls the entire authentication device 100. The CPU 101 reads control programs recorded in a ROM (Read Only Memory) 102 and a RAM (Random Access Memory) 103 and executes various processes.
[0015] The ROM 102 is a non-volatile memory that stores programs and other programs and data necessary for control of the present embodiment. The RAM 103 is a volatile memory and is used as a temporary storage area such as a main memory and a work area.
[0016] The secondary storage device 104 is a rewritable secondary storage device such as a hard disk drive or a flash memory, and stores image information, an image processing program, various setting contents, and the like. These pieces of information are transferred to the RAM 103 and the CPU 101 executes programs and uses data.
[0017] The imaging device (imaging unit) 105 is composed of an imaging lens, an imaging sensor such as a CCD or a CMOS, a video signal processing unit, and the like, and captures video and images. The input device (input unit) 106 is a keyboard, a mouse, or the like, and enables input from a user. The display device (display unit) 107 is a cathode ray tube (CRT) or a liquid crystal display, and displays processing results and the like to the user.
[0018] The network I / F 108 is a modem, a LAN, or the like that connects to a network such as the Internet or an intranet. The bus 109 connects these and performs input / output of data to and from each other.
[0019] The lighting device (lighting unit) 110 is an LED lighting, a fluorescent lamp, or the like, and functions as a device that irradiates light to the user. The blower device (blower unit) 111 is a fan, a blower, or the like, and functions as a device that blows air to the user. The sound output device (sound output unit) 112 is a speaker, and functions as a device that emits a beep sound or a voice message to the user.
[0020] The authentication device 100 in this embodiment reads software that implements the processing corresponding to each step of the flowcharts (FIGS. 3, 4, 5, and 6) described later from the secondary storage device 104, the RAM 103, etc., and executes it using the CPU 101.
[0021] FIG. 2 is a diagram showing the functional configuration of the authentication device 100 in this embodiment.
[0022] The video acquisition unit 201 includes the imaging device 105 and acquires a video by imaging. The face authentication unit 202 performs face authentication on the person in the video acquired by the video acquisition unit 201. The face authentication unit (person authentication unit) 202 in this embodiment includes a face detection unit 203, a face feature extraction unit 204, a face feature matching unit 205, and a first storage unit (registered face feature storage unit) 206.
[0023] The face detection unit 203 detects the face area of a person from the video acquired by the video acquisition unit 201. The face feature extraction unit 204 extracts the feature amount of the face from the face area detected by the face detection unit 203. The face feature matching unit 205 matches with the feature amount stored in the first storage unit 206. The first storage unit 206 corresponds to the RAM 103 and the secondary storage device 104, and stores the feature amount (registered face feature amount) extracted in advance from the person registered as the person to be authenticated together with the identification information of the person.
[0024] The impersonation determination unit 207 determines impersonation in the authentication target person (determines whether the impersonation act is being performed) based on the detection result of the body reaction detection unit 209 described later. Impersonation is, for example, using a photo of a registered person obtained in advance or a still image / moving image displayed on a tablet and holding it in front of the authentication camera, or wearing a 3D mask to impersonate the registered person. The impersonation determination unit 207 in this embodiment includes a body reaction stimulus presentation unit 208, a body reaction detection unit 209, a body reaction evaluation unit 210, and a second storage unit (body reaction evaluation history storage unit) 211.
[0025] The physical reaction stimulation presentation unit 208 selects a stimulation for inducing an unconscious physical reaction of the person to be authenticated (user) from a plurality of stimulation types (kinds), and presents the selected stimulation to the person to be authenticated. The physical reaction detection unit 209 detects an unconscious physical reaction of the person to be authenticated expected in response to the stimulation of the type selected by the physical reaction stimulation presentation unit 208. The physical reaction evaluation unit 210 evaluates the physical reaction of the person to be authenticated detected by the physical reaction detection unit 209 for spoofing determination.
[0026] The second storage unit 211 corresponds to the RAM 103 and the secondary storage device 104, and stores the result of the physical reaction evaluation unit 210 (the history of physical reaction evaluations) as a history together with the ID of the person to be authenticated. The authentication determination unit 212 determines the authentication availability of the person to be authenticated from the authentication result of the face authentication unit 202 and the spoofing determination result of the spoofing determination unit 207. The display unit 213 corresponds to the display device 107, and displays information such as the authentication result output from the authentication determination unit 212.
[0027] Next, an outline of the operation of the authentication device 100 of the present embodiment will be described with reference to FIG. 3. FIG. 3 is a flowchart showing the processing flow of the authentication device 100 according to the present embodiment. By prefixing each step (step) with S, the notation of the step (step) is omitted.
[0028] In S301, the face authentication unit 202 performs a person authentication process. In the present embodiment, face authentication of a person (person to be authenticated) in the video acquired by the video acquisition unit 201 is performed. The face authentication unit 202 outputs the authentication result in the face authentication to the authentication determination unit 212.
[0029] In S302, the authentication determination unit 212 determines whether the person to be authenticated is a registered person based on the authentication result output from the face authentication unit 202. If the person to be authenticated is a registered person, the person to be authenticated who underwent person authentication in S301 is determined as the registered person, and the process proceeds to S303. On the other hand, if the person to be authenticated is not a registered person, the process proceeds to S306 without determining the person to be authenticated who underwent person authentication in S301 as the registered person.
[0030] In S302, the impersonation determination unit 207 performs an impersonation determination on the authentication target person determined as the registered person in S302. The impersonation determination unit 207 outputs the determination result in the impersonation determination process to the authentication approval determination unit 212.
[0031] In S304, the authentication approval determination unit 212 determines whether it is an impersonation (whether the authentication target person is performing an impersonation act) based on the determination result output from the impersonation determination unit 207. If it is not an impersonation, the process proceeds to S305. On the other hand, if it is an impersonation, the process proceeds to S306.
[0032] In S305, the authentication approval determination unit 212 approves the authentication of the authentication target person as "approved (successful)". Further, the authentication approval determination unit 212 displays a message indicating that the authentication of the authentication target person is approved, such as "Authentication succeeded", on the screen of a display device such as the display unit 213. Then the process ends.
[0033] In S306, the authentication approval determination unit 212 disapproves the authentication of the authentication target person as "disapproved (failed)". Further, the authentication approval determination unit 212 displays a message indicating that the authentication of the authentication target person is disapproved, such as "Authentication failed", on the screen of a display device such as the display unit 213. Then the process ends.
[0034] Note that in this embodiment, the result of authentication approval is displayed on the screen of the display unit 213, but this is just an example. For example, it may be interlocked with a gate opening / closing system so that the gate is unlocked when the authentication is approved, or when operating as part of an electronic payment system, it may proceed to the payment procedure, etc. Also, a staff member may check the message of authentication approval or disapproval displayed on the screen and permit or reject the passage of the authentication target person.
[0035] Next, the details of the person authentication (face authentication) process will be described with reference to FIG. 4. FIG. 4 is a flowchart showing the flow of the face authentication process in the authentication apparatus 100 according to the present embodiment. Each step is denoted by prefixing "S" at the beginning, and the notation of the steps is omitted.
[0036] In S401, the video acquisition unit 201 acquires video from the imaging device 105 in units of frame images.
[0037] In S402, the face detection unit 203 detects the face (area) of the person to be authenticated existing in the frame image acquired by the video acquisition unit 201. As a specific method of face detection, for example, a method using a Convolutional Neural Network (CNN) can be mentioned. Note that it is only necessary that the face area can be detected, and it is not limited to this. The face area is represented by the x coordinates and y coordinates of two points, the upper left and the lower right, of a rectangle surrounding the person, with the upper left of the frame image as the origin.
[0038] In S403, the face feature extraction unit 204 extracts face features from the face area. As a specific method of face feature extraction, for example, using a CNN, a large number of combinations of two images of the person himself / herself and one image of another person are prepared in advance, and learning is performed so that the distance between the images of the person himself / herself is close and the distance from the image of another person is far. Note that it is only necessary that the face features necessary for identifying the person can be extracted, and it is not limited to this.
[0039] In S404, the face feature matching unit 205 reads out the face features of all the persons registered in advance from the first storage unit 206. In the first storage unit 206, feature amounts extracted from face images of persons permitted for authentication prepared in advance are stored together with IDs for identifying the persons. Note that since the same processing as in S403 may be performed for feature extraction, the description is omitted.
[0040] In S405, the face feature matching unit 205 performs a comparison between the face features of a plurality of registered persons read by the face feature matching unit 205 in S404 and the face features of the face region included in the image acquired by the video acquisition unit 201 in S401. First, the distance between the feature amounts is calculated for the comparison. Specific methods for calculating the distance include Euclidean distance, L1 distance, cosine distance, etc. The closer the distance is to 0, the higher the likelihood of being the actual person. Therefore, a predetermined threshold is set, and the person ID with a distance below the threshold is determined to be the actual person.
[0041] In S406, the face feature matching unit 205 determines whether the distance between the feature amounts is below a predetermined threshold. If the distance between the feature amounts is below the predetermined threshold, that is, if the distance is below the threshold for the face features of any person ID in the first storage unit 206, it is assumed that the person to be authenticated has been determined to be a registered person, and the process proceeds to S407. On the other hand, if the distance between the feature amounts exceeds the predetermined threshold, it is assumed that the person to be authenticated could not be determined to be a registered person, and the process proceeds to S408.
[0042] In S407, the face feature matching unit 205 outputs the matching result (the result of being able to determine as a registered person) obtained in the matching in S405 to the authentication approval determination unit 212. Then the subsequent process ends. In S408, the face feature matching unit 205 outputs the matching result (the result of not being able to determine as a registered person) obtained in the matching in S405 to the authentication approval determination unit 212. Further, a special person ID (for example, 0 or -1, etc.) is output to the authentication approval determination unit 212 as information indicating that the person could not be determined to be a registered person. Then the subsequent process ends.
[0043] Next, the details of the forgery determination process will be described with reference to FIGS. 5 and 6. FIGS. 5 and 6 are flowcharts showing the processing flow of the forgery determination unit 207 in the authentication device 100 according to the present embodiment. By prefixing each step with S, the notation of the steps is omitted.
[0044] In S501, the physical reaction stimulus presentation unit 208 refers to the second storage unit 211 and reads out the history corresponding to the person ID of the authentication result by the face authentication unit 202. For the sake of simplicity in explanation, first, the case where there is no history corresponding to the person ID (the case of the first authentication) will be described, and then the case where there is a history will be described.
[0045] In S502, the physical reaction stimulus presentation unit 208 selects the stimulus to be presented to the person to be authenticated. The physical reaction stimulus presentation unit 208 stores in the RAM 103, the secondary storage device 104, etc. as a table such as that shown in FIG. 7 the type of stimulus, intensity, expected unconscious physical reaction, and expected reaction time. FIG. 7 is a table showing an example of the content of the stimulus that induces an unconscious physical reaction in the authentication device 100 according to the present embodiment. The physical reaction stimulus presentation unit 208 refers to this table and determines which stimulus (to present) to use for the person to be authenticated.
[0046] In the present embodiment, the physical reaction stimulus presentation unit 208 changes the stimulus selected from the table shown in FIG. 7 according to the number of times of stimulus presentation and the number of times of authentication execution. Specifically, in the case of the first authentication (authentication process), the physical reaction stimulus presentation unit 208 selects the stimulus with the weakest intensity. Then, as the number of times of stimulus presentation and the number of times of authentication execution increase, the physical reaction stimulus presentation unit 208 selects a stimulus with a stronger intensity than the intensity of the previously presented stimulus. That is, the physical reaction stimulus presentation unit 208 selects a stimulus corresponding to the number of times of stimulus presentation and the number of times of authentication execution from among a plurality of different types of stimuli stored in the table.
[0047] Taking the table shown in FIG. 7 as an example, the intensity of the stimulus ranges from 1 to 5. In the present embodiment, it is assumed that the stimulus with an intensity of 1 has the weakest intensity and the stimulus with an intensity of 5 has the strongest intensity. Therefore, in this example, the "display of a mark" with an intensity of 1 is selected by the physical reaction stimulus presentation unit 208. Here, when there are a plurality of the same intensities (when stored in the table), the physical reaction stimulus presentation unit 208 randomly selects from among the stimuli having the same intensity.
[0048] In S503, the physical reaction stimulus presentation unit 208 determines the time (timing) to present the stimulus selected in S502. Since the stimulus in this embodiment is one that elicits an unconscious physical reaction of a person, in S503, the physical reaction stimulus presentation unit 208 determines the time to present the stimulus to the authentication target person at a timing that is difficult to predict. That is, the timing of stimulus presentation is set to a timing that is difficult for the authentication target person to anticipate. For example, it is desirable that the presentation of the stimulus is not synchronized with, for example, the switching of the guidance displayed on the screen during the authentication procedure, and in some cases, the stimulus is presented so as to interrupt.
[0049] In this embodiment, it is assumed that the time determined by the physical reaction stimulus presentation unit 208 is a time randomly determined each time. That is, the physical reaction stimulus presentation unit 208 changes the start time when presenting the stimulus every time authentication is executed. However, it is preferable to set a limit such that the time determined does not cause the user to wait for a long time, within a predetermined time range.
[0050] In S504, when the time (start time) determined in S503 arrives, the physical reaction stimulus presentation unit 208 presents the stimulus selected in S502 to the authentication target person.
[0051] Here, as the content of the stimulus to be presented to the authentication target person, in the example of the table in FIG. 7, for "display of a mark", a mark unrelated to the authentication procedure (for example, a dog, a cat, a symbol, a figure, an arrow, an icon, etc.) is displayed at an arbitrary position on the screen by the display unit 213. Also, for "shining light", the lighting device 110 is used to shine light on the user's face. Also, for "blowing wind", the blower device 111 is used to blow wind on the user's face. Also, for "producing sound", the sound output device 112 is used to emit a beep sound, a voice message, etc. Also, for "displaying a message", a message unrelated to the authentication procedure, for example, news retrieved from the Internet, is displayed on a display device such as the display unit 213.
[0052] These are just examples. In the authentication device 100 of the present embodiment, by inducing the user's unconscious physical reaction, it is confirmed that the face being authenticated is the real face rather than the face displayed on the smartphone in a video or the like. Therefore, other methods may be used as long as they can induce the unconscious physical reaction as described above.
[0053] In S505, the video acquisition unit 201 acquires a video (camera video) from the imaging device 105 in units of frame images.
[0054] In S506, the physical reaction detection unit 209 detects the physical reaction of the person to be authenticated in response to the presentation of the type of stimulus selected by the physical reaction stimulus presentation unit 208 in S502. Here, when detecting the physical reaction of the person to be authenticated, a detection method suitable for detecting the "expected unconscious physical reaction" corresponding to the type of stimulus selected by the physical reaction stimulus presentation unit 208 by the physical reaction detection unit 209 is selected. Specific examples will be described below.
[0055] "Looking up the face movement" and "shaking the head movement" detect the face orientation. As a method for detecting the face orientation, for example, a method using CNN is used to detect facial organs and convert them into positions in three-dimensional space to detect the face orientation. Also, a method of aligning a standard three-dimensional face model with the converted coordinates and obtaining the face orientation from the orientation of the standard three-dimensional face model may be used.
[0056] "Facial expression change" can be determined, for example, by using a method using CNN to detect facial organs and the degree of matching with the standard organ positions of pre-determined facial expressions (such as smiling, anger, etc.).
[0057] "Opening the eyes movement" can be determined, for example, by using a method using CNN to detect the organ points around the eyes and determining the degree of eye opening from the positional relationship of the coordinates of each organ point.
[0058] The "change in pupil area" can be obtained, for example, by using a method that uses a CNN to identify the eye region from organ points around the eyes, detecting the pupil region by performing threshold processing using the luminance difference between the pupil and the iris, and calculating the area ratio to the iris.
[0059] Note that these are just examples, and other methods may be used as long as they can detect physical reactions. Also, in the above examples, all used face images as input, but depending on the physical reaction, upper body or full body images may also be used. Further, for example, depth information may be used by further equipping with a stereo camera, or a method of analyzing voice by equipping with a microphone may be used.
[0060] In S507, the physical reaction detection unit 209 determines whether the "expected unconscious physical reaction" corresponding to the type of stimulus selected by the physical reaction stimulus presentation unit 208 has been detected. If the expected unconscious physical reaction has not been detected, the process proceeds to S508. On the other hand, if the expected unconscious physical reaction has been detected, the process proceeds to S509.
[0061] In S508, the physical reaction detection unit 209 determines whether the expected reaction time has elapsed. If the expected reaction time has elapsed, the process proceeds to S509. On the other hand, if the expected reaction time has not elapsed, the process returns to S505 and the same processing as above is repeated. The expected reaction time is the time when a physical reaction is expected to be seen from the presentation of the stimulus. That is, in the case of a genuine face, a physical reaction should return within this time. Therefore, if the expected reaction time has not elapsed, the process returns to S505 as described above and the detection of the physical reaction is repeated.
[0062] In S509, the body reaction evaluation unit 210 determines whether the body reaction detected in S506 matches the "expected unconscious body reaction" corresponding to the stimulus type selected by the body reaction stimulus presentation unit 208. If the body reaction detected in S506 matches the expected unconscious body reaction (if the expected reaction is detected), the process proceeds to S510. On the other hand, if the body reaction detected in S506 does not match the expected unconscious body reaction (if the expected reaction is not detected), the process proceeds to S511.
[0063] In S510, the body reaction evaluation unit 210 determines whether a body reaction that matches the expected unconscious body reaction has been detected within the expected reaction time. If a body reaction that matches the expected unconscious body reaction has been detected within the expected reaction time, the process proceeds to S512. In this case, since the body reaction detected in S506 matches the expected unconscious body reaction shown in the table of FIG. 7 and the body reaction has been detected within the expected reaction time, the body reaction evaluation unit 210 determines that it is a true face (not a disguise). On the other hand, if a body reaction that matches the expected unconscious body reaction has not been detected within the expected reaction time, the process proceeds to S511. Thus, in the present embodiment, in the processes of S509 and S510, the body reaction evaluation unit 210 evaluates the body reaction detection result.
[0064] In S511, the body reaction evaluation unit 210 determines whether the number of trials of body reaction detection has reached a predetermined number set in advance. If the number of trials of body reaction detection has not reached the predetermined number set in advance, a stronger stimulus is presented to the person to be collated and the detection of the body reaction is attempted again, so the process returns to S502.
[0065] Here, when returning to S502 via the process of S511, the physical reaction stimulus presentation unit 208 refers to the table in FIG. 7 and selects a stimulus of a type with a higher intensity than the previous stimulus presentation. That is, when a physical reaction corresponding to the type of stimulus presented to the person to be authenticated is not detected and the number of detections of the physical reaction has not reached a predetermined number, a stimulus with a higher intensity than the presented stimulus is selected, and the selected stimulus is presented to the person to be authenticated. Also, when a physical reaction corresponding to the type of stimulus presented to the person to be authenticated is detected, but the physical reaction is not detected within the expected reaction time and the number of detections of the physical reaction has not reached a predetermined number, the same applies as above. That is, the physical reaction stimulus presentation unit 208 selects a stimulus with a higher intensity than the presented stimulus and presents the selected stimulus to the person to be authenticated.
[0066] On the other hand, when the number of trials of physical reaction detection reaches a predetermined number set in advance, the process proceeds to S512. The physical reaction evaluation unit 210 determines that it is an impersonation when the number of trials of physical reaction detection reaches a predetermined number set in advance.
[0067] In this way, the physical reaction evaluation unit 210 determines whether the person to be verified is performing an impersonation based on the type of the detected physical reaction of the person to be verified and the time from the presentation of the stimulus until the physical reaction is detected. Note that the number of times as the threshold value in S511 can be arbitrarily set and changed.
[0068] When the expected reaction cannot be detected or, even if it is detected, it is not within the expected reaction time, there is a possibility of impersonation. However, since the consciousness of the person to be recognized may be temporarily distracted or there may be a type of stimulus that is difficult for a person to show a reaction, determining based on a single physical reaction may reduce the accuracy of the impersonation determination. To prevent this, the accuracy of the impersonation determination can be improved by performing the process of S511 described above.
[0069] In S512, the body reaction evaluation unit 210 outputs the spoofing determination result to the authentication approval determination unit 212. In S513, the body reaction evaluation unit 210 stores the history of body reaction evaluations together with the person ID of the authentication result in the second storage unit 211. Then the process ends.
[0070] FIG. 8 is a table showing an example of the evaluation history of spoofing determination in the authentication device 100 according to the present embodiment. FIG. 9 is a table showing another example of the evaluation history of spoofing determination in the authentication device 100 according to the present embodiment.
[0071] Associated with the person ID is, as shown by an example in the table of FIG. 8, the combination of the type of stimulus and the detection result of the body reaction arranged in the order in which they were tried. The example of the table in FIG. 8 shows success in the body detection reaction in the second trial of the type "message display" (indicating that the expected body reaction was detected within the expected reaction time). Therefore, it shows the case where it is evaluated as the real face. Also, since all the other examples shown in the table of FIG. 9 failed in all body detection reactions, it shows the case where it is determined as spoofing.
[0072] In the above, in S501, the case where there is no body reaction evaluation history corresponding to the person ID, which is the processing result in the person authentication in S301 (the first authentication of the person to be authenticated), has been described. The case where there is a body reaction evaluation history will be described below with reference to the examples of the tables shown in FIG. 8 and FIG. 9. In the process of S501 in this case, the examples of the tables in FIG. 8 and FIG. 9 will be used for explanation based on the person ID.
[0073] In S501, if the evaluation history read by the physical reaction stimulus presentation unit 208 from the second storage unit 211 is as shown in the table of FIG. 8, although the previous authentication was successful, the first "mark display" was a failure. Therefore, with this stimulus, it is possible that the user (the person to be authenticated) may have difficulty reacting. Thus, in S502, a stimulus with a higher intensity than the successful second "message display" is selected. Note that if the same stimulus is presented many times, the user may become less likely to react, so it is preferable to select a different type of stimulus from the previous one. That is, when presenting a stimulus to the person to be authenticated again with reference to the second storage unit 211, it is desirable that the physical reaction stimulus presentation unit 208 does not select the stimulus that was selected and presented during the previous authentication. In other words, when determining the authentication eligibility again for a person who has already received a determination of authentication eligibility, it is desirable to select a stimulus different from the stimulus presented during the previous authentication execution with reference to the second storage unit 211. Note that the physical reaction stimulus presentation unit 208 may be configured not to select a stimulus that has already been selected when presenting a stimulus. Here, in the example of the table of FIG. 7, since there is no type with the same intensity as the "message display", a "sound output" with a higher intensity is selected by the physical reaction stimulus presentation unit 208.
[0074] On the other hand, in S501, if the evaluation history read by the physical reaction stimulus presentation unit 208 from the second storage unit 211 is as shown in the table of FIG. 9, since all detections of physical reactions failed in the previous authentication, it is determined that there is an impersonation. Although it is conceivable that the same person may impersonate this time, since this person already knows the stimuli such as "mark display", "message display", and "sound output", countermeasures such as preparing a video suitable for these reactions in advance may be taken. Thus, in such a case, in S502, stimuli of the types in the evaluation history are avoided, and a selection is made from other stimuli. In the example of the table of FIG. 7, the physical reaction stimulus presentation unit 208 selects "shining light".
[0075] By utilizing the evaluation history as described above, it is possible to reduce the number of trials by avoiding stimuli of types that are less likely to elicit a response from a person who has successfully authenticated previously. As a result, spoofing determination can be made in a short time. On the other hand, in the case where it is determined that there was spoofing during the previous authentication, since countermeasures against the person who carried out the spoofing can be rendered ineffective, highly accurate spoofing determination can be performed.
[0076] As described above, according to the authentication device 100 of the present embodiment, it becomes possible to induce an unconscious physical reaction of the user and determine spoofing based on whether the expected physical reaction occurs within the expected time. Thereby, it becomes possible to perform spoofing determination that is robust against sophisticated spoofing acts such as repeatedly presenting a video including a plurality of motion patterns, and does not impose a burden on the user to pay attention for a long time. Therefore, it is possible to provide the authentication device 100 capable of suppressing false determination of spoofing in person authentication.
[0077] In the above-described embodiment, all functions are incorporated into one device. However, for example, the video acquired from the imaging device may be transmitted via a network, and one or both of the face authentication unit 202 and the spoofing determination unit 207 may be arranged on the cloud.
[0078] The present invention can also be realized by supplying a program that realizes one or more functions of the above-described embodiment to a system or device via a network or a storage medium, and having one or more processors in a computer of the system or device read and execute the program. Further, it can also be realized by a circuit (for example, ASIC) that realizes one or more functions.
[0079] As described above, the preferred embodiments of the present invention have been described in detail. However, the present invention is not limited to such specific embodiments, and various modifications and changes are possible within the scope of the gist of the present invention described in the claims.
[0080] The disclosure of the present embodiment includes the following configurations, methods, and programs.
[0081] (Configuration 1) An authentication device for authenticating a person, comprising: Presentation means for presenting a predetermined stimulus that induces an unconscious physical reaction of the person to be authenticated to the person to be authenticated; Detection means for detecting the physical reaction of the person to be authenticated after presenting the stimulus; Determination means for determining whether the person to be authenticated is performing an impersonation act based on the physical reaction detected by the detection means; Determination means for determining the acceptability of authentication of the person to be authenticated based on the determination result of the determination means. An authentication device characterized by the above. An authentication device characterized by the above.
[0082] (Configuration 2) Further comprising face authentication means for performing face authentication of the person to be authenticated in the video, The authentication device according to Configuration 1, wherein the determination means determines the acceptability of authentication of the person to be authenticated based on the authentication result of the face authentication means and the determination result of the determination means.
[0083] (Configuration 3) Further comprising selection means for selecting the stimulus to be presented to the person to be authenticated, The authentication device according to Configuration 1 or 2, wherein the selection means changes the stimulus selected according to the number of times of presenting the stimulus.
[0084] (Configuration 4) The authentication device according to Configuration 3, wherein the selection means selects the stimulus from a table in which a plurality of different types of stimuli are stored based on the number of times of presenting the stimulus.
[0085] (Configuration 5) The authentication device according to any one of Configurations 1 to 4, wherein the presentation means presents the stimulus to the person to be authenticated at a timing that is difficult to predict.
[0086] (Configuration 6) The authentication device according to any one of Configurations 1 to 5, wherein the presentation means changes the start time when presenting the stimulus to the person to be authenticated for each presentation of the stimulus.
[0087] (Configuration 7) The presentation means displays a predetermined mark at an arbitrary position on the screen of the display means as the stimulus, The detection means detects at least one of a face looking-up movement or an eye-opening movement of the person to be authenticated as a physical reaction after the predetermined mark is displayed. The authentication device according to any one of Configurations 1 to 6.
[0088] (Configuration 8) The presentation means irradiates light on the face of the person to be authenticated as the stimulus, The detection means detects a change in the pupil area of the person to be authenticated as a physical reaction after the light is irradiated. The authentication device according to any one of Configurations 1 to 7.
[0089] (Configuration 9) The presentation means blows wind on the face of the person to be authenticated as the stimulus, The detection means detects an eye-opening movement of the person to be authenticated as a physical reaction after the wind is blown. The authentication device according to any one of Configurations 1 to 8.
[0090] (Configuration 10) The presentation means outputs sound as the stimulus, The detection means detects at least one of a head-shaking movement or an eye-opening movement of the person to be authenticated as a physical reaction after the sound is output. The authentication device according to any one of Configurations 1 to 9.
[0091] (Configuration 11) The presentation means displays a message on the screen of the display means as the stimulus, The detection means is configured to detect a change in the expression of the person to be authenticated as a physical reaction after displaying the message, and is the authentication device according to any one of Configurations 1 to 10.
[0092] (Configuration 12) The determination means determines whether the person to be authenticated is performing an impersonation act based on the type of the detected physical reaction of the person to be authenticated and the time from the presentation of the stimulus until the physical reaction is detected, and is the authentication device according to any one of Configurations 1 to 11.
[0093] (Configuration 13) When a physical reaction corresponding to the type of the stimulus presented to the person to be authenticated is detected and the physical reaction is detected within a predetermined time, the determination means determines that the person to be authenticated is not performing an impersonation act, and is the authentication device according to any one of Configurations 1 to 12.
[0094] (Configuration 14) When a physical reaction corresponding to the type of the stimulus presented to the person to be authenticated is not detected and the number of detections of the physical reaction has not reached a predetermined number, the presentation means selects a stimulus with a higher intensity than the presented stimulus and presents the selected stimulus to the person to be authenticated, and is the authentication device according to any one of Configurations 1 to 13.
[0095] (Configuration 15) When a physical reaction corresponding to the type of the stimulus presented to the person to be authenticated is detected but the physical reaction is not detected within a predetermined time and the number of detections of the physical reaction has not reached a predetermined number, the presentation means selects a stimulus with a higher intensity than the presented stimulus and presents the selected stimulus to the person to be authenticated, and is the authentication device according to Configuration 1.
[0096] (Configuration 16) The determination means is configured to determine that the person to be authenticated is performing an impersonation act when the number of detections of the physical reaction has reached a predetermined number, according to any one of Configurations 1 to 14 of the authentication device described above.
[0097] (Configuration 17) The authentication device further includes a storage means for storing, as a history, the type of stimulus presented by the presentation means to the person to be authenticated and the detection result of the detection means together with the identification information of the person. When determining the approval or disapproval of authentication again for a person to be authenticated who has once received a determination of approval or disapproval of authentication, the presentation means refers to the storage means and presents a stimulus different from the stimulus presented at the previous authentication execution, according to any one of Configurations 1 to 16 of the authentication device described above.
[0098] (Configuration 18) The authentication device according to any one of Configurations 1 to 17, characterized in that it has a display control means for displaying the authentication result determined by the determination means on the screen of the display means.
[0099] (Configuration 19) A control method for an authentication device that authenticates a person, comprising: a presentation step of presenting a predetermined stimulus that induces an unconscious physical reaction of the person to be authenticated to the person to be authenticated; a detection step of detecting the physical reaction of the person to be authenticated after presenting the stimulus; a determination step of determining whether the person to be authenticated is performing an impersonation act based on the physical reaction detected in the detection step; a determination step of determining the approval or disapproval of authentication of the person to be authenticated based on the determination result in the determination step. The authentication device is characterized by the above.
[0100] (Configuration 20) A program for operating a computer as the authentication device according to any one of Configurations 1 to 18. for.
Explanation of Signs
[0101] 202 Facial recognition unit 207 Impersonation determination unit 208 Physical reaction stimulation unit 209 Physical reaction detection unit 210 Physical reaction evaluation unit 211 Second storage unit 212 Authentication approval determination unit 213 Display unit
Claims
**Claim 1** An authentication device for authenticating a person, comprising: a presentation means for presenting a predetermined stimulus to the person to be authenticated, which induces an unconscious physical reaction of the person to be authenticated; a detection means for detecting the physical reaction of the person to be authenticated after presenting the stimulus; a determination means for determining whether the person to be authenticated is performing an impersonation act based on the physical reaction detected by the detection means; a decision means for determining whether or not to authenticate the person to be authenticated based on the determination result of the determination means. An authentication device characterized by the above. **Claim 2** further comprising a face authentication means for performing face authentication of the person to be authenticated in the video; The authentication device according to claim 1, wherein the decision means determines whether or not to authenticate the person to be authenticated based on the authentication result of the face authentication means and the determination result of the determination means. **Claim 3** further comprising a selection means for selecting the stimulus to be presented to the person to be authenticated; The authentication device according to claim 1, wherein the selection means changes the stimulus to be selected according to the number of times of presenting the stimulus. **Claim 4** The authentication device according to claim 3, wherein the selection means selects the stimulus from a table in which a plurality of different types of stimuli are stored based on the number of times of presenting the stimulus. **Claim 5** The authentication device according to claim 1, wherein the presentation means presents the stimulus to the person to be authenticated at a timing that is difficult to predict. **Claim 6** The authentication device according to claim 1, wherein the presentation means changes the start time when presenting the stimulus to the person to be authenticated for each presentation of the stimulus. **Claim 7** The presentation means displays a predetermined mark at an arbitrary position on the screen of the display means as the stimulus; The authentication device according to claim 1, wherein the detection means detects at least one of an upward movement of the face or an opening movement of the eyes of the person to be authenticated as a physical reaction after displaying the predetermined mark. **Claim 8** The presentation means irradiates light on the face of the person to be authenticated as the stimulus; The authentication device according to claim 1, wherein the detection means detects a change in the area of the pupil of the person to be authenticated as a physical reaction after irradiating the light. **Claim 9** The presentation means blows wind on the face of the person to be authenticated as the stimulus; The authentication device according to claim 1, wherein the detection means detects an opening movement of the eyes of the person to be authenticated as a physical reaction after blowing the wind.
10. The prompting means outputs sound as the stimulus, and the detecting means detects, as a physical reaction, at least one of a head shaking motion or an eye opening motion of the person to be authenticated after the sound is output. The authentication device according to claim 1, characterized in that.
11. The prompting means displays a message on the screen of the display means as the stimulus, and the detecting means detects a change in the expression of the person to be authenticated as a physical reaction after the message is displayed. The authentication device according to claim 1, characterized in that.
12. Based on the type of the physical reaction of the person to be authenticated detected and the time from the presentation of the stimulus until the physical reaction is detected, the determination means determines whether the person to be authenticated is performing an impersonation act. The authentication device according to claim 1, characterized in that.
13. When a physical reaction corresponding to the type of stimulus presented to the person to be authenticated is detected and the physical reaction is detected within a predetermined time, the determination means determines that the person to be authenticated is not performing an impersonation act. The authentication device according to claim 1, characterized in that.
14. When a physical reaction corresponding to the type of stimulus presented to the person to be authenticated is not detected and the number of detections of the physical reaction has not reached a predetermined number, the prompting means selects a stimulus with a higher intensity than the previously presented stimulus and presents the selected stimulus to the person to be authenticated. The authentication device according to claim 1, characterized in that.
15. When a physical reaction corresponding to the type of stimulus presented to the person to be authenticated is detected but the physical reaction is not detected within a predetermined time and the number of detections of the physical reaction has not reached a predetermined number, the prompting means selects a stimulus with a higher intensity than the previously presented stimulus and presents the selected stimulus to the person to be authenticated. The authentication device according to claim 1, characterized in that.
16. When the number of detections of the physical reaction has reached a predetermined number, the determination means determines that the person to be authenticated is performing an impersonation act. The authentication device according to claim 1, characterized in that.
17. The prompting means further has a storage means for storing, as a history, the type of stimulus presented to the person to be authenticated and the detection result of the detection means together with the identification information of the person. The authentication device according to claim 1, wherein the presentation means refers to the storage means when determining whether to authenticate again a person to be authenticated who has once received a determination on whether authentication is possible, and presents a stimulus different from the stimulus presented at the time of the previous authentication execution.
18. The authentication device according to claim 1, further comprising display control means for displaying, on a screen of a display means, the result of authentication determined by the determination means.
19. A control method for an authentication device that authenticates a person, comprising: a presentation step of presenting a predetermined stimulus that induces an unconscious physical reaction of the person to be authenticated to the person to be authenticated; a detection step of detecting a physical reaction of the person to be authenticated after presenting the stimulus; a determination step of determining whether the person to be authenticated is committing an impersonation act based on the physical reaction detected in the detection step; and a determination step of determining whether to authenticate the person to be authenticated based on the determination result in the determination step. A control method for an authentication device, characterized by the above.
20. A program for operating a computer as the authentication device according to any one of claims 1 to 18 therefor.
Citation Information
Patent Citations
Device for detecting impersonation
JP2008015800A
Impersonation detection device, impersonation detection method, and recording medium
WO2016059786A1