Information processing device, information processing method, and recording medium

Abstract

An information processing device according to the disclosure includes an acquisition unit, a determination unit, and a judgment unit. The acquisition unit acquires situation information that indicates at least one of a surroundings situation, which is the situation of people is the surroundings of a person being judged, and a subject situation, which is the situation of the person being judged. The determination unit determines, on the basis of the situation information, a judgment threshold for judging whether the person being judged is impersonating someone else. On the basis of the determination threshold, the judgment unit judges whether the person being judged is impersonating someone else.

Description

Information Processing Apparatus, Information Processing Method, and Recording Medium

[0001] This disclosure relates to an information processing apparatus, an information processing method, and a recording medium.

[0002] A technique related to this disclosure is disclosed in Patent Document 1. Patent Document 1 discloses a technique related to impersonation determination in face authentication.

[0003] When there is another person near the person to be authenticated and that person is a trustworthy person, the impersonation determination for the person to be authenticated is aborted. More specifically, when it is determined that a trustworthy person existing near the person to be authenticated is paying attention to the person to be authenticated, the impersonation determination for the person to be authenticated is aborted. In such a situation where the person to be authenticated is less likely to perform an impersonation act, the impersonation determination for the person to be authenticated is aborted. In an example of impersonation determination, the user needs to perform a predetermined movement according to an instruction from the system. By aborting the impersonation determination under predetermined conditions, this technique reduces the burden on the user required for the impersonation determination.

[0004] Incidentally, based on the face orientation and line-of-sight direction of a trustworthy person existing near the person to be authenticated, this technique determines whether the person to be authenticated is included in the field of view of that trustworthy person. Then, based on whether the person to be authenticated is included in the field of view of that trustworthy person, this technique determines whether that trustworthy person is paying attention to the person to be authenticated.

[0005] Japanese Unexamined Patent Application Publication No. 2022 - 155061

[0006] In recent years, personal authentication by computer processing has been performed in various scenarios. In personal authentication, the act of impersonating others becomes a problem.

[0007] An example of the purpose of this disclosure is to develop a technique for impersonation determination to determine whether an act of impersonating others is being performed.

[0008] According to one aspect of this disclosure, an information processing device is provided, comprising: acquisition means for acquiring situational information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged; determination means for determining a determination threshold for determining impersonation of the person to be judged based on the situational information; and determination means for performing impersonation of the person to be judged based on the determination threshold.

[0009] Furthermore, according to one aspect of this disclosure, an information processing method is provided in which one or more computers acquire situational information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged; determine a judgment threshold for the impersonation judgment of the person to be judged based on the situational information; and perform the impersonation judgment of the person to be judged based on the judgment threshold.

[0010] Furthermore, according to one aspect of this disclosure, a recording medium is provided which records a program that causes a computer to execute: an acquisition step of acquiring situational information indicating at least one of the surrounding situation of people around the person to be judged and the person's own situation; a determination step of determining a determination threshold for determining whether the person to be judged is impersonating someone based on the situational information; and a determination step of performing an impersonation determination on the person to be judged based on the determination threshold.

[0011] Figure 1 is a diagram showing an example of a functional block diagram of an information processing device. Figure 2 is a flowchart showing an example of the processing flow of an information processing device. Figure 3 is a diagram illustrating an example of the processing of an information processing device. Figure 4 is a diagram showing an example of the hardware configuration of an information processing device. Figure 5 is a flowchart showing another example of the processing flow of an information processing device. Figure 6 is a diagram showing an example of the information processed by an information processing device. Figure 7 is a flowchart showing yet another example of the processing flow of an information processing device.

[0012] The embodiments of this disclosure will be described below with reference to the drawings. In this disclosure, the drawings are associated with one or more embodiments. In all drawings, similar components are denoted by the same reference numerals, and their descriptions are omitted where appropriate.

[0013] <<First Embodiment>> Figure 1 is a functional block diagram showing an overview of the information processing device 10. Figure 2 is a flowchart showing an example of the processing flow executed by the information processing device 10.

[0014] As shown in Figure 1, the information processing device 10 has an acquisition unit 11, a determination unit 12, and a judgment unit 13. These functional units execute the processes shown in the flowchart of Figure 2.

[0015] In S10, the acquisition unit 11 acquires situation information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged. In S11, the determination unit 12 determines the judgment threshold for the impersonation judgment of the person to be judged based on the situation information acquired in S10. In S12, the determination unit 13 performs the impersonation judgment of the person to be judged based on the judgment threshold determined in S11.

[0016] Thus, the information processing device 10 has the characteristic of determining a judgment threshold for impersonation of the person to be judged based on at least one of the surrounding circumstances of the person to be judged and the circumstances of the person to be judged, and performing an impersonation judgment of the person to be judged based on that judgment threshold. Impersonation judgment is the determination of whether the person to be judged is engaging in impersonation.

[0017] When performing impersonation detection based on thresholds, it is desirable to set appropriate thresholds. Setting strict thresholds makes it easier to detect all individuals who have engaged in impersonation behavior. However, in this case, there is a higher possibility of falsely identifying individuals who have not actually engaged in impersonation behavior as having done so. Some kind of verification is required for individuals who have been falsely identified as having engaged in impersonation behavior, but as the number of falsely identified individuals increases, the burden of this verification process increases, which can lead to problems such as a decrease in throughput.

[0018] Furthermore, setting a looser threshold can mitigate the problem of incorrectly identifying individuals who have not actually engaged in impersonation as having done so. As a result, the aforementioned issues such as reduced throughput can be mitigated. However, in this case, the likelihood of incorrectly identifying individuals who have engaged in impersonation as not having done so increases. An increase in such misidentifications is undesirable from a security standpoint.

[0019] In impersonation detection, achieving both high accuracy and high throughput is difficult. To address this problem, the information processing device 10 can determine a threshold for each person to be judged based on at least one of the surrounding circumstances of the person to be judged and the circumstances of the person to be judged. With such an information processing device 10, an appropriate threshold can be set for each person to be judged based on at least one of the surrounding circumstances of the person to be judged and the circumstances of the person to be judged. For example, the information processing device 10 can set a stricter threshold in situations where high accuracy is required, and looser the threshold in situations where such high accuracy is not required to increase throughput.

[0020] For example, if the circumstances surrounding the person being assessed make it unlikely that they would engage in impersonation, the threshold can be lowered. Conversely, if the circumstances surrounding the person being assessed make it highly likely that they would engage in impersonation, the threshold can be lowered. For example, if there are monitors around the person being assessed, or if the monitors are looking at the person being assessed, these are considered situations where the likelihood of impersonation is low. Conversely, if there are no monitors around the person being assessed, or if the monitors are not looking at the person being assessed, these are considered situations where the likelihood of impersonation is high.

[0021] Furthermore, if the circumstances surrounding the person being assessed indicate the possibility that the person is engaging in impersonation, the threshold can be made stricter. Conversely, if the circumstances surrounding the person being assessed do not indicate the possibility that the person is engaging in impersonation, the threshold can be made looser. For example, if the attention of those around the person being assessed is focused on them, it is possible that the person is taking some kind of action to attract the attention of those around them. Therefore, a situation in which the attention of those around the person being assessed is considered to indicate the possibility that the person is engaging in impersonation. On the other hand, a situation in which the attention of those around the person being assessed is not focused on them is considered to not indicate the possibility that the person is engaging in impersonation.

[0022] Furthermore, if the circumstances of the person being assessed indicate that they are engaging in or are likely to engage in impersonation, the threshold can be made stricter. Conversely, if the circumstances of the person being assessed do not indicate that they are engaging in or are likely to engage in impersonation, the threshold can be made looser. For example, a person being assessed who is constantly looking around cautiously or repeatedly looking at a nearby security guard is considered to be engaging in or is likely to engage in impersonation. On the other hand, a person being assessed who does not exhibit such behavior is considered not to be engaging in or is likely to engage in impersonation.

[0023] According to the information processing device 10, for example, an appropriate threshold can be determined for each person to be judged based on at least one of the surrounding circumstances of the person to be judged and the circumstances of the person to be judged themselves.

[0024] Incidentally, in situations where the likelihood of impersonation is low, or where the person being judged is either engaging in or has indicated no possibility of engaging in impersonation, it is conceivable to configure the system to "not perform an impersonation check on the person being judged." However, under no circumstances can the possibility of impersonation being performed be considered "zero." Nevertheless, if the system is configured to "not perform an impersonation check on the person being judged," it will result in the inconvenience of overlooking impersonation that occurs in such situations. This is undesirable from a security standpoint.

[0025] To address this problem, the information processing device 10 with adjustable thresholds allows for a relaxed threshold for determining whether a person is impersonating others when the likelihood of impersonation is low, or when the person being judged is either impersonating others or has not indicated a possibility of doing so. By performing impersonation judgments even in situations where impersonation is unlikely to occur, the information processing device 10 can suppress the inconvenience of overlooking impersonation. Furthermore, by avoiding the use of unnecessarily strict thresholds for impersonation judgments in situations where impersonation is unlikely to occur, throughput can be increased.

[0026] <<Second Embodiment>> <Overview> The information processing device 10 of the second embodiment is a concrete implementation of the configuration of the information processing device 10 of the first embodiment. The information processing device 10 determines a threshold for impersonation detection based on the degree of attention focused on the person being judged by people around that person.

[0027] If the person being judged is engaging in impersonation, as shown in Figure 3, the attention of the people 3 surrounding the person being judged 2 will be concentrated on the person being judged 2. Therefore, the person being judged 2 who is attracting the attention of the people 3 surrounding them is more likely to be engaging in impersonation. On the other hand, the person being judged 2 who is not attracting the attention of the people 3 surrounding them is less likely to be engaging in impersonation. Based on this relationship, the information processing device 10 determines an impersonation detection threshold for each person being judged. This will be explained in detail below.

[0028] <Hardware Configuration> First, an example of the hardware configuration of the information processing device 10 will be described. Each functional unit of the information processing device 10 is realized by any combination of hardware and software. Those skilled in the art will understand that there are various variations in the implementation method and the device. The software includes programs that are pre-installed at the time of shipment of the device, as well as programs downloaded from recording media such as CDs (Compact Discs) or from servers on the Internet.

[0029] Figure 4 is a block diagram illustrating the hardware configuration of the information processing device 10. As shown in Figure 4, the information processing device 10 includes a processor 1A, memory 2A, input / output interface 3A, peripheral circuitry 4A, and bus 5A. The peripheral circuitry 4A includes various modules. The information processing device 10 does not necessarily have peripheral circuitry 4A. The information processing device 10 may also be composed of multiple physically and / or logically separated devices. In this case, each of the multiple devices may have the above hardware configuration.

[0030] Bus 5A is a data transmission path for the processor 1A, memory 2A, peripheral circuit 4A, and input / output interface 3A to send and receive data to and from each other. The processor 1A is a processing unit such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit). Memory 2A is a memory such as RAM (Random Access Memory) or ROM (Read Only Memory). The input / output interface 3A includes interfaces for acquiring information from input devices, external devices, external servers, external sensors, cameras, etc., and interfaces for outputting information to output devices, external devices, external servers, etc. The input / output interface 3A also includes interfaces for connecting to a communication network such as the Internet. Input devices include, for example, a keyboard, mouse, microphone, physical buttons, touch panel, etc. Output devices include, for example, a display, projection device, speaker, printer, mailer, etc. The processor 1A can issue commands to each module and perform calculations based on their calculation results.

[0031] <Functional Configuration> Next, the functional configuration of the information processing device 10 will be described in detail. Figure 1 is an example of a functional block diagram of the information processing device 10. As shown in the figure, the information processing device 10 has an acquisition unit 11, a determination unit 12, and a judgment unit 13.

[0032] The acquisition unit 11 acquires situational information indicating at least one of the following: the surrounding situation, which is the situation of people around the person being assessed, and the person's own situation, which is the situation of the person being assessed. In the second embodiment, the acquisition unit 11 acquires situational information indicating the surrounding situation. An example of acquiring situational information indicating the person's own situation will be explained in the fourth embodiment.

[0033] "Persons subject to judgment" are those who are subject to the impersonation judgment.

[0034] "Impersonation detection" is the process of determining whether the person being authenticated is engaging in impersonation by pretending to be someone else. In other words, the person being authenticated becomes the person being judged. Impersonation detection is performed at the same time as the authentication process. For example, the authentication process may be performed on a person who has been determined not to be impersonating someone else through impersonation detection. Alternatively, impersonation detection may be performed on a person who has successfully authenticated during the authentication process. Furthermore, authentication and impersonation detection may be performed in parallel.

[0035] The "authentication process" is the process of determining whether the person being authenticated is a legitimate user. The authentication process is achieved by comparing biometric information obtained from the person being authenticated at the time of authentication with reference information pre-registered in the database. If there is reference information that matches the biometric information obtained from the person being authenticated at the time of authentication, authentication is successful. On the other hand, if there is no reference information that matches the biometric information obtained from the person being authenticated at the time of authentication, authentication fails.

[0036] In the database, the biometric information of legitimate users is registered as reference information. Examples of biometric information include, but are not limited to, face information, fingerprint information, iris information, voiceprint information, vein information, auricle information, etc. The act of impersonation is to input the biometric information of others into the system instead of one's own biometric information in such authentication processing.

[0037] The scenes and methods of implementing the authentication process are various. Although an example will be described below, it is not limited to these examples.

[0038] In one example, the authentication process may be performed at the timing of entering a predetermined space. Legitimate users are permitted to enter the predetermined space. In the case of this example, the authentication process may be collectively performed on a plurality of authentication targets existing in the passage space through which the authentication target person will pass before entering the predetermined space. The passage space may be a passage such as a corridor or a staircase, or may be a predetermined space such as an entrance, a room, or a parking lot.

[0039] In this case, the information processing device 10 acquires, in real-time processing, a captured image generated by a camera that captures the passage space. Next, the information processing device 10 detects a person (authentication target) from the captured image and extracts the biometric information of the detected authentication target. Then, the information processing device 10 performs the authentication process by comparing the extracted biometric information with the reference information registered in the database in advance. The biometric information used in this example is preferably face information.

[0040] The camera that captures the passage space may be a surveillance camera fixed at a predetermined position or a camera mounted on a moving body. Examples of the moving body include, but are not limited to, a drone, a robot equipped with an autonomous movement mechanism, etc. The moving body moves within the passage space. The camera has a function of detecting visible light and imaging it. As a modification, the camera may have a function of detecting other electromagnetic waves such as infrared rays and imaging them. The camera may capture a moving image or may capture a still image at a predetermined timing.

[0041] As another example, authentication processing may be performed on one person at a time at a passage spot, which the person must pass through one by one before entering a designated space. The passage spot is a gate or entrance, but is not limited to these. In this case, the information processing device 10 acquires biometric information obtained by a biometric information acquisition device installed at the designated passage spot in real time. The information processing device 10 then performs authentication processing by comparing the acquired biometric information with reference information registered in the database in advance. Examples of biometric information acquisition devices include cameras, fingerprint sensors, microphones, and other sensors, but are not limited to these. Examples of biometric information used in this example include facial information, fingerprint information, iris information, voiceprint information, vein information, auricle information, etc., but are not limited to these.

[0042] As another example, authentication may be performed on the authenticated person operating the terminal device when they log in to a server or system or perform a specified operation. For example, in a designated operating space such as a classroom, examination room, or venue, multiple people may each operate a terminal device to receive a class, exam, or service. Only legitimate users can operate terminal devices in the designated operating space to receive classes, exams, or services. Authentication may be performed in such a situation. Legitimate users are authorized to log in to the server or system and perform the specified operations.

[0043] In this example, the information processing device 10 acquires biometric information obtained by the terminal device operated by the person to be authenticated in real time. The information processing device 10 then performs authentication by comparing the acquired biometric information with reference information pre-registered in the database. The terminal device is equipped with a camera, fingerprint sensor, microphone, and other sensors. Examples of terminal devices include, but are not limited to, personal computers, smartphones, tablet devices, mobile phones, portable game consoles, and wearable devices. Examples of biometric information used in this example include, but are not limited to, facial information, fingerprint information, iris information, voiceprint information, vein information, and auricle information.

[0044] "Situation information indicating the surrounding situation" indicates the situation of people around the person to be judged, who is the target of spoofing determination. As described above, the authentication target person becomes the person to be judged. Therefore, the surrounding situation of the person to be judged can also be said to be the surrounding situation of the authentication target person performing the above-described authentication process.

[0045] The acquisition unit 11 can acquire situation information indicating at least one of the following: Whether there are people around the person to be judged, The number of people around the person to be judged, The distance between the people around the person to be judged and the person to be judged, The attributes of the people around the person to be judged, The postures of the people around the person to be judged, The facial orientations of the people around the person to be judged, The gazes of the people around the person to be judged, The actions of the people around the person to be judged

[0046] The acquisition unit 11 acquires a captured image of the space where the person to be judged is located. Then, the acquisition unit 11 can generate situation information indicating the above-described content by analyzing the captured image. The space where the person to be judged is located is, for example, the passage space, passage spot, operation space, etc. where the above-described authentication target person is located.

[0047] The acquisition unit 11 acquires, in real-time processing, the captured image generated by the camera that captures the person to be judged. The camera that captures the person to be judged may be a surveillance camera fixed at a predetermined position or a camera mounted on a moving body. Examples of the moving body include, but are not limited to, drones and robots equipped with an autonomous movement mechanism. The surveillance camera is installed at a position where it can capture the person to be judged. Also, the moving body moves in the space where the person to be judged exists. The camera has a function of detecting visible light and imaging it. As a modification, the camera may have a function of detecting other electromagnetic waves such as infrared rays and imaging them. The camera may capture a moving image or capture a still image at a predetermined timing. The camera may be a monocular camera or a stereo camera.

[0048] The acquisition unit 11 can generate situation information indicating the above-described content by analyzing the captured image. There are various analysis methods, and widely known technologies can be adopted.

[0049] "The area surrounding the person being judged" may also be defined as being within a predetermined distance from the person being judged. In this case, "people in the area surrounding the person being judged" are defined as people who are within a predetermined distance from the person being judged. Alternatively, "the area surrounding the person being judged" may be defined as the space visible in the photograph that includes the person being judged. In this case, people in the area surrounding the person being judged are defined as people visible in the photograph that includes the person being judged. Hereafter, people in the area surrounding the person being judged as defined in this way may be referred to as "surrounding persons."

[0050] The acquisition unit 11 can analyze the captured image and detect people (surrounding persons) in the vicinity of the person to be judged. If the vicinity of the person to be judged is defined as within a predetermined distance from the person to be judged, it is necessary to estimate the distance between the person to be judged and other people. This distance estimation can be achieved using widely known techniques. An example is shown below, but it is not limited to these examples.

[0051] For example, the acquisition unit 11 may determine the height of the person to be judged in the captured image in terms of the number of pixels. Then, assuming that the height of the person to be judged is a predefined standard height [cm], the acquisition unit 11 may calculate a conversion coefficient [cm / number of pixels] by dividing the standard height [cm] by the number of pixels. Then, the acquisition unit 11 may estimate the distance between the person to be judged and other people by multiplying the number of pixels between the person to be judged and other people by the conversion coefficient.

[0052] In addition, the acquisition unit 11 may calculate the distance between the person to be judged and other people using the captured image containing depth information generated by the stereo camera. Alternatively, the acquisition unit 11 may acquire a depth image containing depth information generated by LiDAR (Light Detection and Ranging) in addition to the captured image generated by the camera. The acquisition unit 11 may then use this depth image to calculate the distance between the person to be judged and other people. For example, the acquisition unit 11 may generate three-dimensional spatial information from the depth image and use this information to calculate the distance between two points to calculate the distance between the person to be judged and other people. In this example, LiDAR is installed in the space where the person to be judged is located. As a modification, LiDAR may be attached to a moving object that moves within the space where the person to be judged is located. The definition of a moving object is as described above.

[0053] In addition, the acquisition unit 11 may use artificial intelligence (AI) to generate a depth image containing depth information from the captured image generated by the monocular camera. The acquisition unit 11 may then use this depth image to calculate the distance between the person to be judged and other people.

[0054] Generative AI can be implemented, for example, by a neural network. A neural network contains multiple artificial neurons, each with synapses connecting them. Each synapse has a weight. When such a neural network receives input, it performs calculations using the weights associated with each synapse and produces an output corresponding to the input. The model representing the connection relationships between neurons and synapses is stored in memory, for example, in software form. Alternatively, the model may be implemented as a dedicated circuit. Similarly, the weights of each synapse are also stored in memory in software form. Alternatively, a circuit representing the weights may be implemented in a dedicated circuit. Note that when constructing generative AI using multiple models, it is not necessarily required that all models be stored in the same memory. There are many different models that use such neural networks. Generative AI can be implemented by adopting and substituting a wide variety of models, such as Transformers, Convolutional Neural Networks (CNNs), and Recurrent Neural Networks (RNNs).

[0055] The acquisition unit 11 can generate situational information indicating "whether or not there are people around the person being judged" based on the detection results of people around the person being judged (surrounding persons). The acquisition unit 11 can also generate situational information indicating "the number of people around the person being judged" by counting the number of detected surrounding persons. Furthermore, the acquisition unit 11 can generate situational information indicating "the distance between the person being judged and the people around the person being judged" by calculating the distance between the detected surrounding persons and the person being judged. This distance calculation can be achieved using the estimation method described above.

[0056] Furthermore, the acquisition unit 11 can generate situational information indicating "the attributes of people around the person being judged" by estimating the attributes of the surrounding person based on the visual characteristics of the detected surrounding person. The attributes include, but are not limited to, gender, age, nationality, and occupation. Occupation can be estimated based on the surrounding person's clothing, hat, belongings, shoes, badges, emblems, etc. For example, the acquisition unit 11 can estimate whether or not a surrounding person is a person who monitors the person being judged (a supervisor). Supervisors wear a predetermined uniform. The uniform includes clothing, hats, shoes, etc. The uniform may also have a predetermined badge or emblem attached. Supervisors may also carry a predetermined object (such as a baton). The acquisition unit 11 may estimate whether or not a surrounding person is a supervisor based on the visual characteristics specific to such supervisors. The acquisition unit 11 may perform the above analysis using an estimation model generated in advance by machine learning. The estimation model may be a generating AI. In addition, the facial information of the supervisor may be registered in the information processing device 10 in advance. Furthermore, the acquisition unit 11 may determine whether or not a person in the vicinity is a security guard by using facial recognition based on pre-registered facial information of security guards.

[0057] Furthermore, the acquisition unit 11 can generate situational information indicating "the posture of people around the person being judged" by estimating the posture of detected surrounding people. The estimation of a person's posture can be achieved using widely known posture estimation techniques (OpenPose, MMPose, etc.). In this case, the acquisition unit 11 can generate information indicating a person's posture as situational information using a human body model that connects multiple feature points of the human body.

[0058] Furthermore, the acquisition unit 11 can generate situational information indicating "the orientation of faces around the person being judged" by estimating the orientation of the faces of detected surrounding people. The estimation of face orientation can be achieved using widely known techniques. For example, the acquisition unit 11 may estimate the orientation of the faces of surrounding people based on which part of the head (face, back of the head, temples, etc.) is visible in the captured image, or the positional relationships of various facial features (eyes, nose, mouth, chin, etc.) visible in the captured image. Alternatively, the acquisition unit 11 may generate a 3D face model of the surrounding people based on the captured image, etc., and use this model to estimate the orientation of the faces of the surrounding people. Alternatively, the acquisition unit 11 may use generated AI to estimate the orientation of the faces of surrounding people.

[0059] Furthermore, the acquisition unit 11 can generate situational information indicating "the gaze of people around the person being judged" by estimating the direction of the gaze of detected surrounding people. The direction of a person's gaze can be realized using widely known gaze estimation techniques. In this case, the acquisition unit 11 can generate situational information indicating the direction of a person's gaze.

[0060] Furthermore, the acquisition unit 11 can generate situational information indicating "the actions of people around the person being judged" by estimating whether the detected surrounding person is performing a predefined action. In this case, the acquisition unit 11 can generate situational information indicating whether the surrounding person is performing a predefined action. The predefined action may be, for example, an action that shows interest in the person being judged. Examples of such actions include, but are not limited to, "taking a picture of the person being judged with a camera."

[0061] The determination unit 12 determines a judgment threshold for determining whether a person is impersonating another person, based on the situation information acquired by the acquisition unit 11. The determination unit 12 determines a judgment threshold adjusted for each of the multiple people to be judged, based on the situation information for each of them. The judgment thresholds for each of the multiple people to be judged may differ from one another. In the second embodiment, the determination unit 12 determines a judgment threshold based on the degree of attention focused on the person to be judged by people around them.

[0062] The "judgment threshold" is a threshold used in impersonation detection to compare with the impersonation probability score calculated for each individual being judged. An example of how the impersonation probability score is calculated will be described later, but the higher the probability of impersonation, the higher the impersonation probability score. Individuals whose impersonation probability score is equal to or above the judgment threshold are judged to be engaging in impersonation. On the other hand, individuals whose impersonation probability score is below the judgment threshold are judged not to be engaging in impersonation.

[0063] If the attention of those around the person being judged is focused on that person, it is possible that the person being judged is taking some action to attract the attention of those around them. Therefore, the higher the level of attention from those around the person being judged, the higher the likelihood that the person being judged is engaging in impersonation. Accordingly, the decision unit 12 determines a lower judgment threshold the higher the level of attention from those around the person being judged. On the other hand, the lower the level of attention from those around the person being judged, the lower the likelihood that the person being judged is engaging in impersonation. Accordingly, the decision unit 12 determines a higher judgment threshold the lower the level of attention from those around the person being judged.

[0064] Here, we will explain an example of a method for determining such a judgment threshold.

[0065] In one example, the determination unit 12 can determine a concentration level item value that indicates the degree of attention focused on the person being judged by those around them, based on the situational information. The determination unit 12 can then calculate a judgment threshold based on the determined concentration level item value and a pre-prepared judgment threshold calculation model. The judgment threshold calculation model takes at least one concentration level item value as input and outputs a judgment threshold. The judgment threshold calculation model may also be a function that calculates a judgment threshold from at least one concentration level item value. For example, the function may calculate the judgment threshold by adding an adjustment value calculated based on at least one concentration level item value to a pre-determined "standard threshold". In this case, the higher the degree of attention focused on the person being judged by those around them, the smaller the adjustment value. Conversely, the lower the degree of attention focused on the person being judged by those around them, the larger the adjustment value.

[0066] In addition, the judgment threshold calculation model may be configured to identify and output a judgment threshold corresponding to the input concentration level item value by referring to a table showing the correspondence between at least one concentration level item value and a judgment threshold.

[0067] The "concentration level item value" is a value that indicates the degree to which surrounding individuals are concentrating their attention on the person being judged, and is a value that is identified from the situational information described above. The determination unit 12 determines the concentration level item value from the situational information described above.

[0068] The concentration level item value may include at least one of the following, for example. However, the concentration level item value is not limited to the following examples. - Whether or not there are any people in the vicinity who are looking at, facing, or moving their body towards the person being assessed - Number of simultaneously - Statistical value of the duration for which each of the multiple people in the vicinity continued to look at, face, or move their body towards the person being assessed - Statistical value of the cumulative duration for which each of the multiple people in the vicinity continued to look at, face, or move their body towards the person being assessed - Statistical value of the number of times each of the multiple people in the vicinity looked at, face, or move their body towards the person being assessed - Cumulative time for which at least one person in the vicinity was looking at, facing, or moving their body towards the person being assessed - Whether or not there are any people in the vicinity who have photographed the person being assessed - Number of people in the vicinity who have photographed the person being assessed simultaneously - Statistical value of the duration for which each of the multiple people in the vicinity continued to photograph the person being assessed - Statistical value of the cumulative duration for which each of the multiple people in the vicinity photographed the person being assessed - Statistical value of the number of times each of the multiple people in the vicinity photographed the person being assessed - Cumulative time for which at least one person in the vicinity photographed the person being assessed

[0069] The concentration level item value, "whether or not there are people in the vicinity who are looking at, facing, or moving their body towards the person being judged," is quantified according to a predetermined rule, such as "yes = 1, no = 0." In this example, the higher the concentration level item value, the lower the judgment threshold.

[0070] For example, the determination unit 12 may determine the concentration level item value based on whether or not there are surrounding people who are looking at, facing, or their bodies towards the person to be judged at a predetermined pinpoint determination timing. This determination timing is, for example, the timing at which the judgment threshold is determined, or a predetermined timing around that. Alternatively, the determination unit 12 may determine the concentration level item value based on whether or not there are surrounding people who are looking at, facing, or their bodies towards the person to be judged within a predetermined determination time period. This determination time period is, for example, the time period from the above determination timing to a predetermined time before.

[0071] The determination of whether a surrounding person is directing their gaze, face, or body towards the person to be judged is achieved by processing situational information using widely known techniques. Situational information indicates the direction of the surrounding person's face, gaze, and posture (body orientation). For example, the determination unit 12 may determine that the surrounding person is directing their gaze, face, or body towards the person to be judged if the person to be judged is located in a first direction from the position of the surrounding person. The first direction may be the direction of the surrounding person's face, gaze, or body as indicated in the situational information. In addition, the first direction may include the direction of the surrounding person's face, gaze, or body as indicated in the situational information, and directions in which the difference from these is within a threshold.

[0072] The "number of surrounding people who directed their gaze, face, or body towards the person being judged" may be the number of surrounding people who directed their gaze, face, or body towards the person being judged at the decision timing. Alternatively, the number of surrounding people who directed their gaze, face, or body towards the person being judged may be the number of surrounding people who directed their gaze, face, or body towards the person being judged within the decision time period. For example, the decision unit 12 may count the number of surrounding people who directed their gaze, face, or body towards the person being judged at a predetermined pinpoint decision timing and determine the result as the concentration level item value. Alternatively, the decision unit 12 may count the number of surrounding people who directed their gaze, face, or body towards the person being judged within a predetermined decision time period and determine the result as the concentration level item value. The decision timing and decision time period are as described above. The larger the concentration level item value, the smaller the judgment threshold becomes.

[0073] The "number of surrounding individuals who simultaneously directed their gaze, face, or body towards the person being judged" may also be the number of surrounding individuals who simultaneously directed their gaze, face, or body towards the person being judged at the time of determination. Alternatively, the number of surrounding individuals who simultaneously directed their gaze, face, or body towards the person being judged may be a statistical value of the number of surrounding individuals who simultaneously directed their gaze, face, or body towards the person being judged within the determination time period. Examples of statistical values ​​include, but are not limited to, the maximum value, minimum value, mode, median, and mean. For example, the determination unit 12 may count the number of surrounding individuals who simultaneously directed their gaze, face, or body towards the person being judged at a predetermined pinpoint determination time and determine the result as the concentration level item value. Alternatively, the determination unit 12 may calculate a statistical value of the number of surrounding individuals who simultaneously directed their gaze, face, or body towards the person being judged within a predetermined determination time period and determine the result as the concentration level item value. The determination timing and determination time period are as described above. The larger the concentration level item value, the smaller the determination threshold becomes.

[0074] The "statistical value of the duration for which each of the surrounding individuals continued to direct their gaze, face, or body towards the person being judged" is the statistical value of the duration for which each surrounding individual continued to direct their gaze, face, or body towards the person being judged within the determination time period. Examples of statistical values ​​include the maximum value, minimum value, mode, median, and mean, but are not limited to these. For example, the determination unit 12 may calculate the statistical value of the duration for which each surrounding individual continued to direct their gaze, face, or body towards the person being judged within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the judgment threshold becomes.

[0075] The "statistical value of the cumulative time that each of the surrounding individuals directed their gaze, face, or body towards the person being judged" is the statistical value of the cumulative time that each surrounding individual directed their gaze, face, or body towards the person being judged within the determination time period. Examples of statistical values ​​include the maximum value, minimum value, mode, median, and mean, but are not limited to these. For example, the determination unit 12 may calculate the statistical value of the cumulative time that each surrounding individual directed their gaze, face, or body towards the person being judged within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the judgment threshold becomes.

[0076] The "statistical value of the number of times each of the surrounding individuals directed their gaze, face, or body towards the person being judged" is the statistical value of the number of times each surrounding individual directed their gaze, face, or body towards the person being judged within the determination time period. Examples of statistical values ​​include the maximum value, minimum value, mode, median, and mean, but are not limited to these. For example, the determination unit 12 may calculate the statistical value of the number of times each surrounding individual directed their gaze, face, or body towards the person being judged within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the judgment threshold becomes.

[0077] "The cumulative time during which at least one person in the vicinity directs their gaze, face, or body towards the person being judged" refers to the cumulative time during which at least one person in the vicinity directs their gaze, face, or body towards the person being judged within the determination time period. For example, the determination unit 12 may calculate the cumulative time during which at least one person in the vicinity directs their gaze, face, or body towards the person being judged within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the judgment threshold becomes.

[0078] The concentration level item value for "whether or not there are people in the vicinity who photographed the subject being judged" is quantified according to a predetermined rule, such as "present = 1, absent = 0". In this example, the higher the concentration level item value, the lower the judgment threshold.

[0079] For example, the determination unit 12 may determine the concentration level item value based on whether or not there are surrounding people who have photographed the subject at the determination timing. Alternatively, the determination unit 12 may determine the concentration level item value based on whether or not there are surrounding people who have photographed the subject within a predetermined determination time period. The determination timing and determination time period are as described above.

[0080] The determination of whether a person in the vicinity is photographing the subject is achieved by processing situational information using widely known techniques. Situational information indicates the posture of the person in the vicinity. For example, the determination unit 12 may determine whether a person in the vicinity is photographing based on whether their posture is a predetermined photographic posture. The determination unit 12 may further determine the type of object the person in the vicinity is holding. The determination unit 12 may then determine that a person in the vicinity is photographing if they are holding a photographic device (camera, smartphone, etc.) and are in a photographic posture.

[0081] Furthermore, the determination unit 12 can analyze the captured image and estimate the shooting direction of the person in the vicinity. Estimation of the shooting direction can be achieved using widely known techniques. For example, the determination unit 12 may estimate the shooting direction as the direction facing the outer surface where the lens of the shooting device is located. Alternatively, the determination unit 12 may estimate the shooting direction as the direction from the head of the person taking the picture toward the shooting device held in their hand.

[0082] Furthermore, the determination unit 12 may determine that a person in the vicinity is taking a photograph, and if the person to be determined is located in a second direction from the position of the person in the vicinity or the photographing device, it may determine that the person in the vicinity is taking a photograph of the person to be determined. The second direction may be the shooting direction described above. In addition, the second direction may include the shooting direction described above, and directions in which the difference between them is within a threshold.

[0083] The "number of people in the vicinity who photographed the person to be judged" may be the number of people in the vicinity who photographed the person to be judged at the decision timing. Alternatively, the number of people in the vicinity who photographed the person to be judged may be the number of people in the vicinity who photographed the person to be judged within the decision time period. For example, the decision unit 12 may count the number of people in the vicinity who photographed the person to be judged at a predetermined pinpoint decision timing and determine the result as the concentration level item value. Alternatively, the decision unit 12 may count the number of people in the vicinity who photographed the person to be judged within a predetermined decision time period and determine the result as the concentration level item value. The decision timing and decision time period are as described above. The larger the concentration level item value, the smaller the judgment threshold becomes.

[0084] The "number of surrounding people who simultaneously photographed the person to be judged" may be the number of surrounding people who simultaneously photographed the person to be judged at the time of determination. Alternatively, the number of surrounding people who simultaneously photographed the person to be judged may be a statistical value of the number of surrounding people who simultaneously photographed the person to be judged within the determination time period. Examples of statistical values ​​include, but are not limited to, the maximum value, minimum value, mode, median, and mean. For example, the determination unit 12 may count the number of surrounding people who simultaneously photographed the person to be judged at a predetermined pinpoint determination time and determine the result as the concentration level item value. Alternatively, the determination unit 12 may calculate a statistical value of the number of surrounding people who simultaneously photographed the person to be judged within a predetermined determination time period and determine the result as the concentration level item value. The determination timing and determination time period are as described above. The larger the concentration level item value, the smaller the determination threshold.

[0085] The "statistical value of the duration for which each of the multiple surrounding individuals continued to photograph the subject" is the statistical value of the duration for which each surrounding individual continued to photograph the subject within the determination time period. Examples of statistical values ​​include the maximum value, minimum value, mode, median, and mean, but are not limited to these. For example, the determination unit 12 may calculate the statistical value of the duration for which each surrounding individual continued to photograph the subject within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the determination threshold becomes.

[0086] The "statistical value of the cumulative time each of the surrounding individuals photographed the subject" is the statistical value of the cumulative time each surrounding individual photographed the subject within the determination time period. Examples of statistical values ​​include the maximum value, minimum value, mode, median, and mean, but are not limited to these. For example, the determination unit 12 may calculate the statistical value of the cumulative time each surrounding individual photographed the subject within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the determination threshold becomes.

[0087] The "statistical value of the number of times each of the surrounding individuals photographed the person to be judged" is the statistical value of the number of times each surrounding individual photographed the person to be judged within the determination time period. Examples of statistical values ​​include the maximum value, minimum value, mode, median, mean, etc., but are not limited to these. For example, the determination unit 12 may calculate the statistical value of the number of times each surrounding individual photographed the person to be judged within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the determination threshold becomes.

[0088] "The cumulative time during which at least one person in the vicinity is photographing the subject of judgment" refers to the cumulative time during which at least one person in the vicinity is photographing the subject of judgment within the determination time period. For example, the determination unit 12 may calculate the cumulative time during which at least one person in the vicinity is photographing the subject of judgment within a predetermined determination time period and determine the result as the concentration level item value. The determination time period is as described above. The larger the concentration level item value, the smaller the judgment threshold becomes.

[0089] The determination unit 13 performs an impersonation determination on the person being judged based on the determination threshold determined by the decision unit 12 for each person being judged.

[0090] Specifically, the determination unit 13 calculates an impersonation probability score for each person being evaluated. The higher the probability of impersonation, the higher the impersonation probability score. The determination unit 13 then makes an impersonation determination for the person being evaluated based on the relationship between the impersonation probability score and the determination threshold. If the impersonation probability score is equal to or greater than the determination threshold, the determination unit 13 determines that the person being evaluated is engaging in impersonation. On the other hand, if the impersonation probability score is less than the determination threshold, the determination unit 13 determines that the person being evaluated is not engaging in impersonation.

[0091] Here, we will explain the process for calculating the impersonation possibility score. The determination unit 13 can calculate the impersonation possibility score using various technologies.

[0092] For example, the determination unit 13 may calculate a spoofing probability score based on the posture of the person being judged. People who engage in spoofing behavior may adopt certain postures. For example, in facial recognition, a person engaging in spoofing behavior may hold a piece of paper or display showing another person's face in their hand and hold it near their own face. The determination unit 13 analyzes the captured image of the person being judged to identify the person's posture, and if the identified posture is a predefined posture for spoofing behavior, it can calculate a higher spoofing probability score. The determination unit 13 may calculate a higher spoofing probability score the greater the similarity between the identified posture and the predefined posture for spoofing behavior.

[0093] Alternatively, the determination unit 13 may calculate a spoofing probability score using widely known biometric detection (liveness detection) technology. In this case, the determination unit 13 can calculate a higher spoofing probability score if it determines that the object is not a living being.

[0094] Furthermore, the person being authenticated, who is the subject of the authentication process described above, also becomes the person being judged for impersonation. Therefore, the impersonation judgment is performed at the same time that the authentication process takes place.

[0095] In one example, authentication and impersonation detection may be performed at the time of entry into a designated space. Legitimate users are permitted to enter the designated space. In this example, impersonation detection may be performed collectively on multiple individuals present in a passage space that the person being detected passes through before entering the designated space. The passage space may be a corridor or stairwell, or an entrance, room, parking lot, or other open space.

[0096] In this case, the information processing device 10 acquires the captured image generated by the camera that photographs the passage space in real time. The information processing device 10 then analyzes the captured image and calculates a score indicating the likelihood of the person being judged in the image being an imposter.

[0097] The camera used to photograph the passage space may be a fixed surveillance camera or a camera mounted on a mobile device. Examples of mobile devices include, but are not limited to, drones and robots equipped with autonomous movement mechanisms. The mobile device moves within the passage space. The camera has the function of detecting visible light and capturing images. As a variation, the camera may also have the function of detecting and capturing other electromagnetic waves such as infrared rays. The camera may capture moving images or still images at predetermined timings. In addition, other sensors such as temperature sensors may be used instead of a camera.

[0098] As another example, a fraud detection check may be performed on a single person to be judged at a passage spot, which is passed through one person at a time before entering a designated space. The passage spot may be a gate or entrance, but is not limited to these. In this case, the information processing device 10 acquires the captured image generated by a camera installed at the designated passage spot photographing the person to be judged in real time. The information processing device 10 then analyzes the acquired captured image to calculate a fraud probability score for the person to be judged as captured in the image. The camera has the function of detecting and imaging visible light. As a modification, the camera may also have the function of detecting and imaging other electromagnetic waves such as infrared rays. The camera may capture moving images or still images at predetermined timings. In addition, other sensors such as temperature sensors may be used instead of the camera.

[0099] As another example, impersonation detection may be performed on the person operating the terminal device when they log in to a server or system or perform a specified operation. For example, in a designated operating space such as a classroom, examination hall, or venue, multiple people may each operate a terminal device to receive a class, exam, or service. Only legitimate users are allowed to operate terminal devices in the designated operating space to receive classes, exams, or services. Impersonation detection may be performed in such a scenario. Legitimate users are permitted to log in to servers and systems and perform specified operations.

[0100] In this example, the information processing device 10 acquires the captured image in real time, which is generated when a terminal device operated by the person being judged takes a picture of the person being judged. The information processing device 10 then analyzes the acquired captured image and calculates a score indicating the likelihood of the person being judged being impersonated as depicted in the image. The camera has the function of detecting and imaging visible light. As a modification, the camera may also have the function of detecting and imaging other electromagnetic waves such as infrared rays. The camera may capture moving images or still images at predetermined timings. In addition, other sensors such as temperature sensors may be used instead of the camera.

[0101] Next, an example of the processing flow of the information processing device 10 will be explained using the flowchart in Figure 5. In this example, the information processing device 10 acquires a captured image of multiple people present in a predetermined space (such as the passage space mentioned above) and performs an impersonation determination on the multiple people included in the captured image. The purpose here is to explain the processing flow. Details of each process have been described above, so explanations will be omitted here as appropriate.

[0102] First, the information processing device 10 acquires the captured image (S20). Next, the information processing device 10 detects a person from the captured image (S21). Then, the information processing device 10 determines that one of the detected people is the person to be judged (S22).

[0103] Next, the information processing device 10 acquires situational information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged (S23). Then, based on the situational information acquired in S23, the information processing device 10 determines the judgment threshold for the impersonation judgment of the person to be judged (S24).

[0104] Next, the information processing device 10 performs an impersonation determination on the person being judged based on the determination threshold determined in S24 (S25). For example, the information processing device 10 calculates an impersonation possibility score for the person being judged and compares the calculated impersonation possibility score with the determination threshold. If the impersonation possibility score is equal to or greater than the determination threshold, the information processing device 10 can determine that the person being judged is engaging in impersonation. On the other hand, if the impersonation possibility score is less than the determination threshold, the information processing device 10 can determine that the person being judged is not engaging in impersonation. The information processing device 10 then outputs the result of the impersonation determination in S25 (S26).

[0105] Next, if there are still people who have not been determined to be subject to judgment (Yes in S27), the information processing device 10 returns to S22 and repeats the process. If there are no people who have not been determined to be subject to judgment (No in S27), and there is no input to terminate the process (No in S28), the information processing device 10 returns to S20 and repeats the process.

[0106] Furthermore, the information processing device 10 may perform authentication processing for individuals who, in the impersonation determination in S25, are determined not to have committed impersonation. The information processing device 10 may also output the result of the authentication processing. In this case, the information processing device 10 does not perform authentication processing for individuals who, in the impersonation determination in S25, are determined to have committed impersonation.

[0107] In addition, the information processing device 10 may first perform an authentication process on the person to be judged determined in S22. Then, if the authentication process is successful, the information processing device 10 may perform the processes S23 to S26 on that person. In this case, the information processing device 10 does not perform the processes S23 to S26 on the person to be judged who failed the authentication process.

[0108] In addition, the information processing device 10 may perform authentication processing and processing S23 to S26 in parallel for the person to be judged as determined in S22.

[0109] The information processing device 10 can manage user management information that shows the results of the impersonation detection and authentication processes described above. Figure 6 schematically shows an example of such user management information.

[0110] The user management information includes multiple items such as detection number, location information, authentication, impersonation detection, and user ID (identifier).

[0111] The "detection number" is information that identifies the individuals detected from the captured images acquired in S20, and it is a number assigned to each of those individuals. Each time a new person is detected from the captured images, a new detection number is assigned to that newly detected person.

[0112] "Location information" refers to information indicating the location of each detected person. Location information may also refer to information indicating the location within the captured image. The information processing device 10 tracks the person detected within the captured image. Based on the results of this tracking, the information processing device 10 can update the location information.

[0113] "Authentication" indicates the result of the authentication process. For individuals who have not undergone authentication, the value of this item will be blank.

[0114] The "Impersonation Detection" field shows the result of the impersonation detection. For individuals who have not undergone impersonation detection, the value for this item will be blank.

[0115] The "User ID" is the unique identifier of each individual, determined through the authentication process.

[0116] The information processing device 10 may output such user management information to the operator. For example, the information processing device 10 may output user management information as shown in Figure 6 via a display or projection device of its own device. In addition, the information processing device 10 may transmit user management information as shown in Figure 6 to an external device.

[0117] In addition, the information processing device 10 may process the captured image based on user management information as shown in Figure 6 and output the processed image. The processed image is an image in which information is superimposed on the captured image. In the processed image, the results of the authentication process and the results of the impersonation detection are displayed, linked to each of the multiple people included in the captured image.

[0118] Based on this output information, the operator can recognize the results of the authentication process and the impersonation detection for each individual.

[0119] <Effects and Effects> The information processing device 10 of the second embodiment can achieve the same effects and effects as the information processing device 10 of the first embodiment.

[0120] Furthermore, the information processing device 10 can determine a judgment threshold based on the degree to which surrounding individuals are focused on the person being judged. If the person being judged is engaging in impersonation, as shown in Figure 3, the attention of the people 3 surrounding the person being judged 2 is likely to be focused on the person being judged 2. Therefore, the person being judged 2 who is attracting the attention of the people 3 surrounding them is more likely to be engaging in impersonation. On the other hand, the person being judged 2 who is not attracting the attention of the people around them is less likely to be engaging in impersonation. Based on this relationship, the information processing device 10 can appropriately determine the impersonation judgment threshold for each person being judged.

[0121] Furthermore, the information processing device 10 can acquire situational information indicating the surrounding circumstances as described above. By processing such characteristic situational information, the information processing device 10 can accurately determine the degree to which people around the person to be judged (surrounding individuals) are focusing their attention on the person to be judged. Based on the degree to which surrounding individuals are focusing their attention on the person to be judged, the information processing device 10 can appropriately determine a judgment threshold. For example, the information processing device 10 can determine a concentration level item value related to the degree to which surrounding individuals are focusing their attention on the person to be judged from the situational information indicating the surrounding circumstances as described above. The information processing device 10 can determine various types of characteristic concentration level item values ​​as described above. Based on such characteristic concentration level item values, the information processing device 10 can appropriately determine a judgment threshold.

[0122] According to this information processing device 10, it is possible to accurately and appropriately determine the degree to which surrounding individuals are focusing their attention on the person being judged, and to appropriately determine a judgment threshold based on the result.

[0123] Furthermore, the information processing device 10 can determine a judgment threshold adjusted for each individual to be judged. The information processing device 10 can then perform impersonation detection by comparing the judgment threshold determined for each individual with the impersonation probability score calculated for each individual. By using a judgment threshold customized for each individual to perform impersonation detection in this way, the information processing device 10 can achieve both high accuracy and high throughput.

[0124] <<Third Embodiment>> In the third embodiment, the information processing device 10 determines a threshold for impersonation detection based on whether the surrounding environment of the person to be judged is conducive to impersonation. If the surrounding environment of the person to be judged is conducive to impersonation, the likelihood of the person to engage in impersonation increases. On the other hand, if the surrounding environment of the person to be judged is not conducive to impersonation, the likelihood of the person to engage in impersonation decreases. Based on this relationship, the information processing device 10 determines a threshold for impersonation detection for each person to be judged. This will be explained in detail below.

[0125] The determination unit 12 determines the judgment threshold for impersonation detection of the person to be judged, based on the status information acquired by the acquisition unit 11. The determination unit 12 determines a judgment threshold adjusted for each of the multiple persons to be judged, based on the status information of each of the multiple persons to be judged. The judgment thresholds for each of the multiple persons to be judged may differ from one another.

[0126] In the third embodiment, the determination unit 12 determines a judgment threshold for impersonation based on whether the surrounding environment of the person to be judged is conducive to impersonation. The judgment threshold is as described in the second embodiment.

[0127] If the surrounding environment of the person being judged is conducive to impersonation, the likelihood of the person being judged committing impersonation increases. Therefore, the determination unit 12 determines a relatively low judgment threshold when the surrounding environment of the person being judged is conducive to impersonation. On the other hand, if the surrounding environment of the person being judged is not conducive to impersonation, the likelihood of the person being judged committing impersonation decreases. Therefore, the determination unit 12 determines a relatively high judgment threshold when the surrounding environment of the person being judged is not conducive to impersonation.

[0128] Here, we will explain an example of a method for determining such a judgment threshold.

[0129] In one example, the determination unit 12 can determine, based on situational information, a surrounding environment item value indicating whether or not the surrounding environment of the person to be judged is conducive to impersonation. The determination unit 12 can then calculate a judgment threshold based on the determined surrounding environment item value and a pre-prepared judgment threshold calculation model. The judgment threshold calculation model takes at least one surrounding environment item value as input and outputs a judgment threshold. The judgment threshold calculation model may also be a function that calculates a judgment threshold from at least one surrounding environment item value. For example, the function may calculate the judgment threshold by adding an "adjustment value calculated based on at least one surrounding environment item value" to a pre-determined "standard threshold". If the surrounding environment of the person to be judged is conducive to impersonation, the adjustment value will be relatively small. If the surrounding environment of the person to be judged is not conducive to impersonation, the adjustment value will be relatively large.

[0130] In addition, the judgment threshold calculation model may be configured to identify and output a judgment threshold corresponding to the input value of at least one ambient environment item by referring to a table showing the correspondence between at least one ambient environment item value and a judgment threshold.

[0131] The "surrounding environment item value" is a value that relates to whether the surrounding environment of the person being judged is conducive to impersonation, and is a value identified from the situation information described above. The determination unit 12 determines the surrounding environment item value from the situation information described above.

[0132] The surrounding environment item values ​​may include, for example, at least one of the following. However, the surrounding environment item values ​​are not limited to the following examples. • Whether or not there are people with a specified attribute around the person being assessed • The number of people with a specified attribute around the person being assessed • The distance between the person being assessed and the people with a specified attribute around the person being assessed • Statistical values ​​of the distance between the person being assessed and each of the multiple people with a specified attribute around the person being assessed • Whether or not people with a specified attribute around the person being assessed are directing their gaze, face, or body towards the person being assessed • The number of people with a specified attribute who are directing their gaze, face, or body towards the person being assessed • The distance between the person being assessed and the people with a specified attribute who are directing their gaze, face, or body towards the person being assessed • Statistical values ​​of the distance between the person being assessed and each of the multiple people with a specified attribute who are directing their gaze, face, or body towards the person being assessed

[0133] A "person with specified attributes" is someone who monitors the person being evaluated. A person with specified attributes might be, for example, a supervisor.

[0134] The surrounding environment item value, "whether or not there are people with specified attributes around the person being judged," is quantified according to a predetermined rule, such as "present = 1, absent = 0." In this example, the larger the surrounding environment item value, the higher the judgment threshold.

[0135] The determination unit 12 can determine whether or not there is a monitor among the surrounding people described in the second embodiment, and can determine the surrounding environment item value based on the result. If there are people with predetermined attributes around the person being judged, it becomes difficult to carry out impersonation. If there are no people with predetermined attributes around the person being judged, it becomes easier to carry out impersonation.

[0136] For example, the determination unit 12 may determine the surrounding environment item value based on whether or not there are people with predetermined attributes around the person to be judged at a predetermined pinpoint determination timing. This determination timing is, for example, the timing at which the judgment threshold is determined, or a predetermined timing around that. Alternatively, the determination unit 12 may determine the concentration level item value based on whether or not there were people with predetermined attributes around the person to be judged within a predetermined determination time period. This determination time period is, for example, the time period from the above determination timing to a predetermined time before.

[0137] The "number of people with a predetermined attribute surrounding the person to be judged" may be the number of people with a predetermined attribute who were around the person to be judged at the time of determination. Alternatively, the number of people with a predetermined attribute surrounding the person to be judged may be the number of people with a predetermined attribute who were around the person to be judged during the determination time period. For example, the determination unit 12 may count the number of people with a predetermined attribute who were around the person to be judged at a predetermined pinpoint determination time and determine the result as the surrounding environment item value. Alternatively, the determination unit 12 may count the number of people with a predetermined attribute who were around the person to be judged during a predetermined determination time period and determine the result as the surrounding environment item value. The determination timing and determination time period are as described above. The more people with a predetermined attribute surrounding the person to be judged there are, the more difficult it becomes to carry out impersonation. The larger the surrounding environment item value, the larger the determination threshold becomes.

[0138] The "distance between the person being judged and a person with a predetermined attribute who is in the vicinity of the person being judged" may be the distance between the person being judged and a person with a predetermined attribute who was in the vicinity of the person being judged at the time of determination. Alternatively, the distance between the person being judged and a person with a predetermined attribute who is in the vicinity of the person being judged may be the distance between the person being judged and a person with a predetermined attribute who was in the vicinity of the person being judged during the determination time period. For example, the determination unit 12 may calculate the distance between the person being judged and a person with a predetermined attribute who was in the vicinity of the person being judged at a predetermined pinpoint determination time, and determine the result as the surrounding environment item value. Alternatively, the determination unit 12 may calculate the distance between the person being judged and a person with a predetermined attribute who was in the vicinity of the person being judged during a predetermined determination time period, and determine the result as the surrounding environment item value. The determination timing and determination time period are as described above. The smaller the distance between the person being judged and a person with a predetermined attribute who is in the vicinity of the person being judged, the more difficult it becomes to carry out impersonation. The smaller the surrounding environment item value, the larger the determination threshold becomes.

[0139] "Statistical values ​​of the distance between each of several people with predetermined attributes surrounding the person to be judged and the person to be judged" may be statistical values ​​of the distance between each of several people with predetermined attributes surrounding the person to be judged and the person to be judged at the timing of the decision. Alternatively, statistical values ​​of the distance between each of several people with predetermined attributes surrounding the person to be judged and the person to be judged may be statistical values ​​of the distance between each of several people with predetermined attributes surrounding the person to be judged and the person to be judged during the decision time period. Examples of statistical values ​​include maximum value, minimum value, mode, median, mean, etc., but are not limited to these. For example, the decision unit 12 may calculate statistical values ​​of the distance between each of several people with predetermined attributes surrounding the person to be judged and the person to be judged at a predetermined pinpoint decision timing and decide the result as the surrounding environment item value. Alternatively, the decision unit 12 may calculate statistical values ​​of the distance between each of several people with predetermined attributes surrounding the person to be judged and the person to be judged during a predetermined decision time period and decide the result as the surrounding environment item value. The decision timing and decision time period are as described above. The smaller the statistical value of the distance between the person being judged and each of several people with specified attributes in their vicinity, the more difficult it becomes to impersonate them. The smaller the value of the surrounding environment item, the higher the judgment threshold.

[0140] The surrounding environment item value, "whether a person with specified attributes in the vicinity of the person being assessed is looking at, facing, or moving their body towards the person being assessed," is quantified according to a predetermined rule, such as "looking = 1, not looking = 0." In this example, the larger the surrounding environment item value, the higher the judgment threshold.

[0141] For example, the determination unit 12 may determine the surrounding environment item value based on whether or not there is a person with a predetermined attribute who is looking at, facing or moving their body towards the person to be judged at a predetermined pinpoint determination timing. Alternatively, the determination unit 12 may determine the surrounding environment item value based on whether or not there is a person with a predetermined attribute who is looking at, facing or moving their body towards the person to be judged within a predetermined determination time period. The determination timing and determination time period are as described above. If a person with a predetermined attribute who is in the vicinity of the person to be judged is looking at, facing or moving their body towards the person to be judged, it becomes difficult to carry out impersonation. If a person with a predetermined attribute who is in the vicinity of the person to be judged is not looking at, facing or moving their body towards the person to be judged, it becomes easier to carry out impersonation.

[0142] The "number of people with a predetermined attribute who directed their gaze, face, or body towards the person to be judged" may also be the number of people with a predetermined attribute who directed their gaze, face, or body towards the person to be judged at the decision timing. Alternatively, the number of people with a predetermined attribute who directed their gaze, face, or body towards the person to be judged may also be the number of people with a predetermined attribute who directed their gaze, face, or body towards the person to be judged within the decision time period. For example, the decision unit 12 may count the number of people with a predetermined attribute who directed their gaze, face, or body towards the person to be judged at a predetermined pinpoint decision timing and determine the result as the surrounding environment item value. Alternatively, the decision unit 12 may count the number of people with a predetermined attribute who directed their gaze, face, or body towards the person to be judged within a predetermined decision time period and determine the result as the surrounding environment item value. The decision timing and decision time period are as described above. The more people with a predetermined attribute who directed their gaze, face, or body towards the person to be judged, the more difficult it becomes to commit impersonation. The larger the surrounding environment item value, the larger the judgment threshold becomes.

[0143] "The distance between the person with a predetermined attribute who is directing their gaze, face, or body toward the person to be judged and the person to be judged" may be the distance between the person with a predetermined attribute who is directing their gaze, face, or body toward the person to be judged and the person to be judged at the decision timing. Alternatively, the distance between the person with a predetermined attribute who is directing their gaze, face, or body toward the person to be judged may be the distance between the person with a predetermined attribute who is directing their gaze, face, or body toward the person to be judged and the person to be judged within the decision time period. For example, the decision unit 12 may calculate the distance between the person with a predetermined attribute who is directing their gaze, face, or body toward the person to be judged and the person to be judged at a predetermined pinpoint decision timing and determine the result as the surrounding environment item value. Alternatively, the decision unit 12 may calculate the distance between the person with a predetermined attribute who is directing their gaze, face, or body toward the person to be judged and the person to be judged within a predetermined decision time period and determine the result as the surrounding environment item value. This distance may be the distance at the time when the person with a predetermined attribute is directing their gaze, face, or body toward the person to be judged, or it may be the distance at any other time. The decision timing and decision time period are as described above. The smaller the distance between the person with the specified attribute who is looking at, facing, or moving towards the person being judged, the more difficult it becomes to impersonate them. The smaller the value of the surrounding environment item, the higher the judgment threshold.

[0144] "Statistical values ​​of the distance between each of several persons with predetermined attributes who are directing their gaze, face, or body toward the person to be judged and the person to be judged" may be statistical values ​​of the distance between each of several persons with predetermined attributes who are directing their gaze, face, or body toward the person to be judged and the person to be judged at the decision timing. Alternatively, statistical values ​​of the distance between each of several persons with predetermined attributes who are directing their gaze, face, or body toward the person to be judged and the person to be judged may be statistical values ​​of the distance between each of several persons with predetermined attributes who are directing their gaze, face, or body toward the person to be judged and the person to be judged within the decision time period. Examples of statistical values ​​include maximum value, minimum value, mode, median, mean, etc., but are not limited to these. For example, the decision unit 12 may calculate statistical values ​​of the distance between each of several persons with predetermined attributes who are directing their gaze, face, or body toward the person to be judged and the person to be judged at a predetermined pinpoint decision timing and determine the result as the surrounding environment item value. Alternatively, the decision unit 12 may calculate statistical values ​​of the distance between each of several persons with predetermined attributes who are directing their gaze, face, or body toward the person to be judged and the person to be judged within a predetermined decision time period and determine the result as the surrounding environment item value. The distance in question may be the distance at the moment when a person with a specified attribute turns their gaze, face, or body towards the person being judged, or it may be the distance at any other time. The determination timing and determination time period are as described above. The smaller the statistical value of the distance between each of the multiple persons with specified attributes who turn their gaze, face, or body towards the person being judged and the person being judged, the more difficult it becomes to impersonate someone. The smaller the value of the surrounding environment item, the higher the judgment threshold.

[0145] Other configurations of the information processing device 10 can be the same as those in the first and second embodiments.

[0146] According to the information processing device 10 of the third embodiment, the same effects and advantages as those of the information processing device 10 of the first and second embodiments can be achieved.

[0147] Furthermore, the information processing device 10 can determine a threshold for impersonation detection based on whether the surrounding environment of the person being judged is conducive to impersonation. If the surrounding environment of the person being judged is conducive to impersonation, the likelihood of the person being judged committing impersonation increases. On the other hand, if the surrounding environment of the person being judged is not conducive to impersonation, the likelihood of the person being judged committing impersonation decreases. Based on this relationship, the information processing device 10 can appropriately determine a threshold for impersonation detection for each person being judged.

[0148] Furthermore, the information processing device 10 can acquire situational information indicating the surrounding circumstances as described above. By processing such characteristic situational information, the information processing device 10 can accurately determine whether the surrounding environment of the person being judged is conducive to impersonation. Based on the result of this determination, the information processing device 10 can appropriately determine a judgment threshold. For example, the information processing device 10 can determine surrounding environment item values ​​related to whether the surrounding environment of the person being judged is conducive to impersonation from the situational information indicating the surrounding circumstances as described above. The information processing device 10 can determine various types of characteristic surrounding environment item values ​​as described above. Based on such characteristic surrounding environment item values, the information processing device 10 can appropriately determine a judgment threshold.

[0149] According to this information processing device 10, it is possible to accurately and appropriately determine whether the surrounding environment of the person to be judged is conducive to impersonation, and to appropriately determine a judgment threshold based on the result.

[0150] <<Fourth Embodiment>> The information processing device 10 of the fourth embodiment determines a judgment threshold for impersonation based on the circumstances of the person being judged. If the circumstances of the person being judged indicate that the person is engaging in or is likely to engage in impersonation, the information processing device 10 lowers the judgment threshold. On the other hand, if the circumstances of the person being judged do not indicate that the person is engaging in or is likely to engage in impersonation, the information processing device 10 raises the judgment threshold. This will be explained in detail below.

[0151] The acquisition unit 11 acquires situational information that indicates the personal situation of the person being assessed.

[0152] "Situational information indicating the person's situation" indicates the situation of the person being judged who is the subject of the impersonation judgment. The acquisition unit 11 can acquire situational information indicating at least one of the following: • The posture of the person being judged • The direction of the face of the person being judged • The gaze of the person being judged • The actions of the person being judged

[0153] The acquisition unit 11 can identify at least one of the posture, face direction, and gaze of the person to be judged by utilizing the technology for detecting the posture, face direction, and gaze of surrounding persons as described in the second embodiment. The acquisition unit 11 can then generate situational information indicating at least one of the posture, face direction, and gaze of the person to be judged.

[0154] Furthermore, the acquisition unit 11 can generate situational information indicating "the actions of the person being judged" by estimating whether or not the person being judged is engaging in a predefined suspicious behavior. In this case, the acquisition unit 11 can generate situational information indicating whether or not the person being judged is engaging in a predefined suspicious behavior.

[0155] Predefined suspicious behaviors are actions that a person who is impersonating someone or is likely to do so may perform. For example, predefined suspicious behaviors may include loitering, looking around nervously, or looking around cautiously. The detection of a person who performs such suspicious behavior can be achieved using widely known technologies (such as image analysis).

[0156] The determination unit 12 determines the judgment threshold for impersonation detection of the person to be judged, based on the status information acquired by the acquisition unit 11. The determination unit 12 determines a judgment threshold adjusted for each of the multiple persons to be judged, based on the status information of each of the multiple persons to be judged. The judgment thresholds for each of the multiple persons to be judged may differ from one another.

[0157] In the fourth embodiment, the determination unit 12 determines the judgment threshold for impersonation based on whether the person's own circumstances indicate that they are or are likely to be engaging in impersonation. The judgment threshold is as described in the second embodiment.

[0158] The determination unit 12 determines a relatively low determination threshold if the personal circumstances of the person being judged indicate that they are engaging in or are likely to engage in impersonation. On the other hand, if the personal circumstances of the person being judged do not indicate that they are engaging in or are likely to engage in impersonation, it determines a relatively high determination threshold.

[0159] Here, we will explain an example of a method for determining such a judgment threshold.

[0160] In one example, the determination unit 12 can determine a personal status item value that indicates whether the personal status of the person being judged is engaging in or is likely to engage in impersonation, based on the situation information. The determination unit 12 can then calculate a judgment threshold based on the determined personal status item value and a pre-prepared judgment threshold calculation model. The judgment threshold calculation model takes at least one personal status item value as input and outputs a judgment threshold. The judgment threshold calculation model may also be a function that calculates a judgment threshold from at least one personal status item value. The function may, for example, calculate the judgment threshold by adding an "adjustment value calculated based on at least one personal status item value" to a pre-determined "standard threshold". If the personal status of the person being judged is engaging in or is likely to engage in impersonation, the adjustment value becomes relatively small. If the personal status of the person being judged is not engaging in or is not likely to engage in impersonation, the adjustment value becomes relatively large.

[0161] In addition, the judgment threshold calculation model may be configured to identify and output a judgment threshold corresponding to the input at least one personal status item value by referring to a table showing the correspondence between at least one personal status item value and a judgment threshold.

[0162] The "personal status item value" is a value that relates to whether the personal status of the person being judged indicates that they are or are likely to be engaging in impersonation, and is a value identified from the status information described above. The determination unit 12 determines the personal status item value from the status information described above.

[0163] The personal status item values ​​may include at least one of the following, for example. However, the personal status item values ​​are not limited to the following examples. • Whether the person being assessed performed a predefined suspicious behavior. • Whether the person being assessed directed their gaze, face, or body towards a designated person. • The duration for which the person being assessed continued to direct their gaze, face, or body towards a designated person. • The cumulative duration for which the person being assessed directed their gaze, face, or body towards a designated person. • The number of times the person being assessed directed their gaze, face, or body towards a designated person. • Whether the person being assessed did not direct their face or body towards a designated person, but directed their gaze. • The duration for which the person being assessed did not direct their face or body towards a designated person, but continued to direct their gaze. • The cumulative duration for which the person being assessed did not direct their face or body towards a designated person, but directed their gaze. • The number of times the person being assessed did not direct their face or body towards a designated person, but directed their gaze.

[0164] "The designated person" is someone who is vigilant when the person being judged engages in impersonation, for example, a person who monitors the person being judged (a supervisor).

[0165] The value for the "personal status item" indicating whether or not the person being judged has engaged in a predefined suspicious behavior is quantified according to a predetermined rule, such as "did = 1, did not = 0". In this example, the larger the value for the personal status item, the smaller the judgment threshold becomes.

[0166] The determination unit 12 can determine the value of the personal status item by referring to the situation information described above. For example, the determination unit 12 may determine the value of the personal status item based on whether or not the person being judged has engaged in suspicious behavior at a predetermined pinpoint determination timing. This determination timing is, for example, the timing at which the judgment threshold is determined, or a predetermined timing around that timing. Alternatively, the determination unit 12 may determine the value of the personal status item based on whether or not the person being judged has engaged in suspicious behavior within a predetermined determination time period. This determination time period is, for example, the time period from the above determination timing to a predetermined time before that time.

[0167] The value for the "whether the person being judged directed their gaze, face, or body towards the designated person" item is quantified according to a predetermined rule, such as "directed = 1, not directed = 0". In this example, the larger the value for this personal status item, the smaller the judgment threshold. A person being judged who directs their gaze, face, or body towards the designated person and observes that person indicates they are engaging in or are likely to engage in impersonation.

[0168] For example, the determination unit 12 may determine the value of the personal status item based on whether or not the person to be judged directed their gaze, face, or body towards a designated person at a predetermined pinpoint determination timing. Alternatively, the determination unit 12 may determine the value of the personal status item based on whether or not the person to be judged directed their gaze, face, or body towards a designated person within a predetermined determination time period. The determination timing and determination time period are as described above.

[0169] The determination of whether the person being judged has directed their gaze, face, or body towards a designated person can be achieved using the same technology as the determination of whether a surrounding person has directed their gaze, face, or body towards the person being judged, as described in the second embodiment.

[0170] "The duration for which the person being judged continued to direct their gaze, face, or body towards the designated person" refers to the duration for which the person being judged continued to direct their gaze, face, or body towards the designated person within the determination time period. If the person being judged directs their gaze, face, or body towards the designated person multiple times within the determination time period, statistical values ​​of those durations may be used. Examples of statistical values ​​include, but are not limited to, the maximum value, minimum value, mode, median, and mean. For example, the determination unit 12 may calculate the duration for which the person being judged continued to direct their gaze, face, or body towards the designated person within a predetermined determination time period and determine the result as the personal status item value. The determination time period is as described above. The larger the personal status item value, the smaller the judgment threshold.

[0171] "The cumulative time that the person being judged directs their gaze, face, or body towards the designated person" is the cumulative time that the person being judged directs their gaze, face, or body towards the designated person within the determination time period. For example, the determination unit 12 may calculate the cumulative time that the person being judged directs their gaze, face, or body towards the designated person within a predetermined determination time period and determine the result as the personal status item value. The determination time period is as described above. The larger the personal status item value, the smaller the judgment threshold becomes.

[0172] "The number of times the person being judged directed their gaze, face, or body towards the designated person" refers to the number of times the person being judged directed their gaze, face, or body towards the designated person within the determination time period. For example, the determination unit 12 may calculate the number of times the person being judged directed their gaze, face, or body towards the designated person within a predetermined determination time period and determine the result as the personal status item value. The determination time period is as described above. The larger the personal status item value, the smaller the judgment threshold becomes.

[0173] The value for the "whether the person being assessed directed their gaze towards the designated person without turning their face or body towards them" item is quantified according to a predetermined rule, such as "directed = 1, not directed = 0". In this example, the larger the value for this item, the smaller the judgment threshold. A situation where a person directs only their gaze towards the designated person without turning their face or body towards them can be considered a situation where they are observing the designated person without being noticed by those around them. A person being assessed who engages in such behavior is either engaging in or potentially engaging in impersonation.

[0174] For example, the determination unit 12 may determine the value of the personal status item based on whether or not the person being judged directed their gaze towards a designated person without turning their face and body towards the designated person at a predetermined pinpoint determination timing. Alternatively, the determination unit 12 may determine the value of the personal status item based on whether or not the person being judged directed their gaze towards a designated person without turning their face and body towards the designated person within a predetermined determination time period. The determination timing and determination time period are as described above.

[0175] The determination of whether the person being judged directed their gaze towards a designated person without turning their face and body towards that person can be achieved using the same technology as the determination of whether a surrounding person directed their gaze, face, or body towards the person being judged, as described in the second embodiment. That is, the determination unit 12 can use the technology described in the second embodiment to determine whether the person being judged turned their face towards the designated person. The determination unit 12 can also use the technology described in the second embodiment to determine whether the person being judged turned their body towards the designated person. The determination unit 12 can also use the technology described in the second embodiment to determine whether the person being judged directed their gaze towards the designated person. Based on these results, the determination unit 12 can determine whether the person being judged directed their gaze towards the designated person without turning their face and body towards that person.

[0176] "The duration for which the person being judged did not turn their face or body towards the designated person, but continued to direct their gaze towards that person" refers to the duration within the determination time period for which the person being judged did not turn their face or body towards the designated person, but continued to direct their gaze towards that person. If the person being judged did not turn their face or body towards the designated person, but continued to direct their gaze towards that person multiple times within the determination time period, statistical values ​​of those durations can be used. Examples of statistical values ​​include, but are not limited to, the maximum value, minimum value, mode, median, and mean. For example, the determination unit 12 may calculate the duration for which the person being judged did not turn their face or body towards the designated person, but continued to direct their gaze towards that person, within a predetermined determination time period, and determine the result as the personal status item value. The determination time period is as described above. The larger the personal status item value, the smaller the judgment threshold becomes.

[0177] "The cumulative time during which the person being assessed does not turn their face or body towards the designated person, but directs their gaze towards them" is the cumulative time during which the person being assessed does not turn their face or body towards the designated person, but directs their gaze towards them. For example, the determination unit 12 may calculate the cumulative time during which the person being assessed does not turn their face or body towards the designated person, but directs their gaze towards them, within a predetermined determination time period, and determine the result as the personal status item value. The determination time period is as described above. The larger the personal status item value, the smaller the determination threshold becomes.

[0178] "The number of times the person being judged did not turn their face or body towards the designated person, but directed their gaze towards them" refers to the number of times the person being judged did not turn their face or body towards the designated person, but directed their gaze towards them, within the determination time period. For example, the determination unit 12 may calculate the number of times the person being judged did not turn their face or body towards the designated person, but directed their gaze towards them, within a predetermined determination time period, and determine the result as the personal status item value. The determination time period is as described above. The larger the personal status item value, the smaller the judgment threshold becomes.

[0179] Other configurations of the information processing device 10 can be the same as those of the first to third embodiments.

[0180] According to the information processing device 10 of the fourth embodiment, the same effects and advantages as those of the information processing device 10 of the first to third embodiments can be achieved.

[0181] Furthermore, the information processing device 10 can determine a judgment threshold for impersonation based on the circumstances of the person being judged. Specifically, if the circumstances of the person being judged indicate that the person is engaging in or is likely to engage in impersonation, the information processing device 10 can lower the judgment threshold. Conversely, if the circumstances of the person being judged do not indicate that the person is engaging in or is likely to engage in impersonation, the information processing device 10 can raise the judgment threshold.

[0182] Furthermore, the information processing device 10 can acquire situational information indicating the person's status as described above. By processing such characteristic situational information, the information processing device 10 can accurately determine whether the person's own status indicates that the person is engaging in or is likely to engage in impersonation. Based on the result of this determination, the information processing device 10 can appropriately determine a determination threshold. For example, the information processing device 10 can determine a personal status item value related to whether the person's own status indicates that they are engaging in or are likely to engage in impersonation, based on the situational information indicating the person's status as described above. The information processing device 10 can determine various types of characteristic personal status item values ​​as described above. Based on such characteristic personal status item values, the information processing device 10 can appropriately determine a determination threshold.

[0183] According to this information processing device 10, it is possible to accurately and appropriately determine whether the situation of the person being judged indicates that they are or are likely to be engaging in impersonation, and to appropriately determine a judgment threshold based on the result.

[0184] <<Fifth Embodiment>> The information processing device 10 of the fifth embodiment determines a judgment threshold by combining several of the methods described in the second to fourth embodiments. This will be explained in detail below.

[0185] The determination unit 12 determines the judgment threshold for impersonation detection of the person to be judged, based on the status information acquired by the acquisition unit 11. The determination unit 12 determines a judgment threshold adjusted for each of the multiple persons to be judged, based on the status information of each of the multiple persons to be judged. The judgment thresholds for each of the multiple persons to be judged may differ from one another.

[0186] In the fifth embodiment, the determination unit 12 determines a judgment threshold by combining several of the methods described in the second to fourth embodiments. The judgment threshold is as described in the second embodiment. An example of a method for determining such a judgment threshold is described below.

[0187] In one example, the determination unit 12 can determine at least two of the following based on the situation information: the concentration level item value described in the second embodiment, the surrounding environment item value described in the third embodiment, and the personal situation item value described in the fourth embodiment. The determination unit 12 can then calculate a judgment threshold based on the determined item values ​​and a pre-prepared judgment threshold calculation model. The determined item values ​​include at least two of the concentration level item value, the surrounding environment item value, and the personal situation item value. Hereinafter, the determined item values ​​will be referred to as "determined item values".

[0188] The judgment threshold calculation model takes at least one decision item value as input and outputs a judgment threshold. The judgment threshold calculation model may also be a function that calculates a judgment threshold from at least one decision item value. The function may, for example, calculate the judgment threshold by adding an adjustment value calculated based on at least one decision item value to a predetermined "reference threshold".

[0189] In addition, the judgment threshold calculation model may be configured to identify and output a judgment threshold corresponding to the input at least one decision item value by referring to a table showing the correspondence between at least one decision item value and the judgment threshold.

[0190] Other configurations of the information processing device 10 can be the same as those of the first to fourth embodiments.

[0191] The information processing device 10 of the fifth embodiment can achieve the same effects as the information processing device 10 of the first to fourth embodiments. Furthermore, the information processing device 10 that determines the judgment threshold by combining multiple methods described in the second to fourth embodiments can determine the judgment threshold more appropriately.

[0192] <<Sixth Embodiment>> In the sixth embodiment, the information processing device 10 first performs an impersonation determination on multiple targets based on a predetermined common threshold common to all targets. Then, the information processing device 10 performs an impersonation determination again on targets that have been determined to have the potential to be impersonating based on the impersonation determination based on the common threshold. In this second impersonation determination, the information processing device 10 applies the techniques described in the first to fifth embodiments. That is, in this second impersonation determination, the information processing device 10 determines a determination threshold for each target and performs the impersonation determination using that determination threshold. This will be explained in detail below.

[0193] The determination unit 13 performs a first determination and a second determination to determine whether the person being determined is impersonating someone else.

[0194] The "first determination" is the initial determination. In the first determination, the determination unit 13 determines whether a subject is impersonating another subject based on a predetermined common threshold common to multiple subjects. The purpose of the first determination is to remove subjects from the processing list who are clearly not impersonating another person.

[0195] The "common threshold" is a threshold used in impersonation detection that is compared with the impersonation probability score calculated for each individual being judged. The method for calculating the impersonation probability score is as described in the second embodiment. As mentioned above, the purpose of the first judgment is to remove individuals who are clearly not being judged for impersonation from the processing target. For this reason, the common threshold is set strictly. For example, the common threshold will be a smaller value than the judgment threshold. In one example, the standard threshold described in the second to fifth embodiments may be used as the common threshold.

[0196] The "second determination" is performed after the first determination. Individuals determined to be potentially impersonating others in the first determination become subject to the second determination. Individuals determined not to be potentially impersonating others in the first determination are excluded from the second determination. In the second determination, the determination unit 13 performs an impersonation determination on each individual based on a determination threshold adjusted for each individual.

[0197] Furthermore, in the first determination, if a person subject to re-evaluation who may be impersonating someone is detected, the acquisition unit 11 may acquire status information of the person subject to re-evaluation. The determination unit 12 may then determine a determination threshold for the person subject to re-evaluation based on the status information acquired in response to the detection. The determination unit 13 may then perform a second determination on the person subject to re-evaluation based on the determination threshold determined in this way.

[0198] The acquisition unit 11 does not need to acquire status information of the person who was determined not to be a suspect in the first determination. The decision unit 12 does not need to determine the determination threshold for the person who was determined to be a suspect. The determination unit 13 does not need to perform a second determination on the person who was determined to be a suspect.

[0199] Next, an example of the processing flow of the information processing device 10 will be explained using the flowchart in Figure 7. In this example, the information processing device 10 acquires a captured image of multiple people present in a predetermined space (such as the passage space mentioned above) and performs an impersonation determination on the multiple people included in the captured image. The purpose here is to explain the processing flow. Details of each process have been described above, so explanations will be omitted here as appropriate.

[0200] First, the information processing device 10 acquires the captured image (S30). Next, the information processing device 10 detects a person from the captured image (S31). Then, the information processing device 10 determines that one of the detected people is the person to be judged (S32).

[0201] Next, the information processing device 10 performs an impersonation determination on the person being judged based on a common threshold common to all persons being judged (S33: First determination). For example, the information processing device 10 calculates an impersonation possibility score for the person being judged and compares the calculated impersonation possibility score with the common threshold. If the impersonation possibility score is equal to or greater than the common threshold, the information processing device 10 can determine that the person being judged may be engaging in impersonation. On the other hand, if the impersonation possibility score is less than the common threshold, the information processing device 10 can determine that the person being judged is not engaging in impersonation.

[0202] If the system determines that the person being judged has not committed impersonation (No. in S34), the information processing device 10 outputs the result of the impersonation judgment in S33 (S35).

[0203] On the other hand, if the system determines that the person being judged may be engaging in impersonation (Yes in S34), the information processing device 10 proceeds to S36. In S36, the information processing device 10 acquires situational information indicating at least one of the surrounding circumstances, which are the circumstances of people around the person being judged, and the person's own circumstances, which are the circumstances of the person being judged. Based on the situational information acquired in S36, the information processing device 10 then determines the judgment threshold for the impersonation judgment of the person being judged (S37).

[0204] Next, the information processing device 10 performs an impersonation determination on the person being judged based on the determination threshold determined in S37 (S38: second determination). For example, the information processing device 10 calculates an impersonation possibility score for the person being judged and compares the calculated impersonation possibility score with the determination threshold. If the impersonation possibility score is equal to or greater than the determination threshold, the information processing device 10 can determine that the person being judged is engaging in impersonation. On the other hand, if the impersonation possibility score is less than the determination threshold, the information processing device 10 can determine that the person being judged is not engaging in impersonation. The information processing device 10 may also use the impersonation possibility score calculated in the first determination in S33 in the second determination in S38. In addition, the information processing device 10 may calculate an impersonation possibility score in S38 separately from the calculation of the impersonation possibility score in S33. The information processing device 10 then outputs the result of the impersonation determination in S38 (S35).

[0205] Next, if there are still people who have not been determined to be subject to judgment (Yes in S39), the information processing device 10 returns to S32 and repeats the process. If there are no people who have not been determined to be subject to judgment (No in S39), and there is no input to terminate the process (No in S40), the information processing device 10 returns to S30 and repeats the process.

[0206] Furthermore, the information processing device 10 may perform authentication processing for individuals who, in the impersonation determination (first determination) in S34, are determined not to have committed impersonation. Also, the information processing device 10 may perform authentication processing for individuals who, in the impersonation determination (second determination) in S38, are determined not to have committed impersonation. The information processing device 10 may also output the results of the authentication processing. In this case, the information processing device 10 does not perform authentication processing for individuals who, in the impersonation determination (second determination) in S38, are determined to have committed impersonation.

[0207] In addition, the information processing device 10 may first perform an authentication process on the person to be judged determined in S32. Then, if the authentication process is successful, the information processing device 10 may perform the processes S33 to S38 on that person. In this case, the information processing device 10 does not perform the processes S33 to S38 on the person to whom the authentication process was unsuccessful.

[0208] In addition, the information processing device 10 may perform the authentication process and the processes in S33 to S38 in parallel for the person to be judged as determined in S32.

[0209] Other configurations of the information processing device 10 can be the same as those of the first to fifth embodiments.

[0210] The information processing device 10 of the sixth embodiment can achieve the same effects as the information processing device 10 of the first to fifth embodiments. Furthermore, the information processing device 10 can perform a two-stage impersonation determination. Specifically, the information processing device 10 first performs a first determination with a strict threshold, and removes individuals who are clearly not impersonating from the processing target. Then, the information processing device 10 can perform a second determination on individuals who were not removed in the first determination. In the second determination, the technology described in the first to fifth embodiments is employed, and an impersonation determination is performed by determining a determination threshold for each individual.

[0211] With this information processing device 10, the processing described in the first to fifth embodiments only needs to be performed for some of the individuals to be judged. In other words, with this information processing device 10, the acquisition of situational information and the determination of judgment thresholds only need to be performed for some of the individuals to be judged. The acquisition of situational information and the determination of judgment thresholds can be avoided for other individuals to be judged. With this information processing device 10, the processing load on the computer can be reduced.

[0212] <<Modified Version>> <Modified Version 1> The determination unit 12 may calculate the concentration level item value by the following method in addition to or instead of the concentration level item value described in the second embodiment. In this modified version as well, the same effects and advantages as in the first to sixth embodiments are achieved.

[0213] For example, the determination unit 12 may calculate the concentration level item value based on a predetermined function and situation information. The following formula (1) is an example of such a function.

[0214]

[0215] p ０ p is the probability of paying attention to impersonation behavior. １ This represents the probability of focusing on non-impersonation behavior. Non-impersonation behavior is any behavior other than impersonation behavior. １ is, p ０ It is a value smaller than [X]. X is a variable that indicates the degree of focus of gaze.

[0216] Λ represents the probability p of paying attention to impersonation behavior. ０ Likelihood function L(p) based on ０ / X) and "the probability p of focusing attention on non-impersonation behavior" １ Likelihood function L(p) based on １ This is the likelihood ratio with ( / X).

[0217] Note that the function used to calculate the concentration level value is not limited to the example above. For example, the function used to calculate the concentration level value may be a function that increases monotonically with respect to "the number of people in the vicinity who directed their gaze, face, or body towards the person being judged."

[0218] In addition, the function used to calculate the concentration level value may be a function that decreases monotonically with respect to "the number of people in the vicinity who are not directing their gaze, face, or body towards the person being judged."

[0219] Alternatively, the function used to calculate the concentration level value may be a function that increases monotonically with respect to the difference between "the number of people in the vicinity who are directing their gaze, face, or body towards the person being judged" and "the number of people in the vicinity who are not directing their gaze, face, or body towards the person being judged."

[0220] Alternatively, the function used to calculate the concentration level value may be a function that decreases monotonically with respect to the difference between "the number of people in the vicinity who are not looking at the target person with their gaze, face, or body" and "the number of people in the vicinity who are looking at the target person with their gaze, face, or body."

[0221] In these examples, the determination unit 12 counts "the number of people in the vicinity who are not directing their gaze, face, or body towards the person being judged" based on the situational information. The determination unit 12 can then use the result of this count to calculate the concentration level item value.

[0222] <Modification 2> The determination unit 12 may determine at least one of the following concentration level item values ​​in addition to or instead of the concentration level item values ​​described in the second embodiment and the concentration level item values ​​of Modification 1: • Whether or not there are surrounding persons who have directed their gaze, face, or body toward the person to be determined for a certain period of time or longer. • The number of surrounding persons who have directed their gaze, face, or body toward the person to be determined for a certain period of time or longer. • Whether or not there are surrounding persons whose cumulative time of directing their gaze, face, or body toward the person to be determined is a certain period of time or longer. • The number of surrounding persons whose cumulative time of directing their gaze, face, or body toward the person to be determined is a certain period of time or longer. • Whether or not there are surrounding persons who have continuously photographed the person to be determined for a certain period of time or longer. • The number of surrounding persons who have continuously photographed the person to be determined for a certain period of time or longer. • Whether or not there are surrounding persons whose cumulative time of photographing the person to be determined is a certain period of time or longer. • The number of surrounding persons whose cumulative time of photographing the person to be determined is a certain period of time or longer.

[0223] The determination unit 12 can determine the concentration level item value by detecting surrounding persons who meet the above conditions within the determination time period and counting their number, in the same manner as the determination of the concentration level item value described in the second embodiment. The definition of the determination time period is the same as in the first to sixth embodiments. The "certain time" is predetermined. In this modified example, the same effects and advantages as in the first to sixth embodiments are achieved.

[0224] Although this disclosure has been described above with reference to embodiments, this disclosure is not limited to the embodiments described above. Various modifications to the structure and details of this disclosure are possible, which can be understood by those skilled in the art within the scope of this disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0225] Furthermore, the flowchart used in the above explanation shows multiple steps (processes) in sequence. However, the execution order of the steps performed in each embodiment is not limited to the order in which they are described. In each embodiment, the order of the illustrated steps can be changed to the extent that it does not impede the content.

[0226] Some or all of the above embodiments may also be described as follows, but are not limited to the following: 1. An information processing apparatus comprising: an acquisition means for acquiring situation information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's situation, which is the situation of the person to be judged; a determination means for determining a determination threshold in the impersonation determination of the person to be judged based on the situation information; and a determination means for performing the impersonation determination of the person to be judged based on the determination threshold. 2. The information processing apparatus according to claim 1, characterized in that the acquisition means acquires situation information indicating at least one of the following: whether or not there are people around the person to be judged, the number of people around the person to be judged, the distance between the people around the person to be judged and the person to be judged, the attributes of the people around the person to be judged, the posture of the people around the person to be judged, the direction of the faces of the people around the person to be judged, the gaze of the people around the person to be judged, and the actions of the people around the person to be judged. 3. The information processing apparatus according to claim 1 or 2, characterized in that the acquisition means acquires situational information indicating at least one of the following: the posture of the person to be judged, the direction of the face of the person to be judged, the gaze of the person to be judged, and the actions of the person to be judged. 4. The information processing apparatus according to any one of claims 1 to 3, characterized in that the determination means determines the determination threshold based on the degree of attention focused on the person to be judged by surrounding persons who are people in the vicinity of the person to be judged.5. The information processing device according to 4, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not there is a surrounding person who has turned their gaze, face or body towards the person to be determined; the number of surrounding people who have turned their gaze, face or body towards the person to be determined; the number of surrounding people who have turned their gaze, face or body towards the person to be determined at the same time; a statistical value of the duration for which each of the multiple surrounding people has continued to turn their gaze, face or body towards the person to be determined; a statistical value of the cumulative time for which each of the multiple surrounding people has turned their gaze, face or body towards the person to be determined; a statistical value of the number of times each of the multiple surrounding people has turned their gaze, face or body towards the person to be determined; and the cumulative time for which at least one of the surrounding people has turned their gaze, face or body towards the person to be determined. 6. The information processing apparatus according to 4 or 5, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not there are any surrounding persons who have photographed the person to be determined; the number of surrounding persons who have photographed the person to be determined; the number of surrounding persons who have photographed the person to be determined simultaneously; a statistical value of the duration for which each of the multiple surrounding persons continued to photograph the person to be determined; a statistical value of the cumulative time for which each of the multiple surrounding persons photographed the person to be determined; a statistical value of the number of times each of the multiple surrounding persons photographed the person to be determined; and the cumulative time for which at least one of the surrounding persons has photographed the person to be determined. 7. The information processing apparatus according to any one of 1 to 6, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not there are any persons with predetermined attributes in the vicinity of the person to be determined; the number of persons with predetermined attributes in the vicinity of the person to be determined; the distance between the persons with predetermined attributes in the vicinity of the person to be determined and the person to be determined; and a statistical value of the distance between each of the multiple persons with predetermined attributes in the vicinity of the person to be determined and the person to be determined.8. The information processing device according to any one of 1 to 7, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not a person of a predetermined attribute in the vicinity of the person to be judged is directing their gaze, face, or body towards the person to be judged; the number of persons of the predetermined attribute who are directing their gaze, face, or body towards the person to be judged; the distance between the person of the predetermined attribute who is directing their gaze, face, or body towards the person to be judged and the person to be judged; and statistical values ​​of the distance between each of the multiple persons of the predetermined attribute who are directing their gaze, face, or body towards the person to be judged and the person to be judged. 9. The information processing device according to 7 or 8, characterized in that the person of the predetermined attribute is a person who is monitoring the person to be judged. 10. The information processing device according to any one of 1 to 9, characterized in that the determination means determines the determination threshold based on whether or not the person to be judged has performed a predefined suspicious behavior. 11. The information processing apparatus according to any one of 1 to 10, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not the person to be judged directed their gaze, face or body towards a predetermined person; the duration for which the person to be judged continued to direct their gaze, face or body towards the predetermined person; the cumulative duration for which the person to be judged directed their gaze, face or body towards the predetermined person; and the number of times the person to be judged directed their gaze, face or body towards the predetermined person. 12. The information processing apparatus according to any one of 1 to 10, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not the person to be judged directed their gaze towards a predetermined person without directing their face and body towards the predetermined person; the duration for which the person to be judged continued to direct their gaze towards a predetermined person without directing their face and body towards the predetermined person; the cumulative duration for which the person to be judged directed their gaze towards a predetermined person without directing their face and body towards the predetermined person; and the number of times the person to be judged directed their gaze towards a predetermined person without directing their face and body towards the predetermined person. 13. The information processing apparatus according to 11 or 12, characterized in that the predetermined person is a person monitoring the person to be judged. 14. The information processing apparatus according to any one of 1 to 13, characterized in that the determination means determines the determination threshold adjusted for each of the multiple persons to be determined, based on the status information of each of the persons to be determined.15. The information processing apparatus according to any one of 1 to 14, characterized in that the determination means performs a first determination to determine whether a target person is impersonating someone based on a predetermined common threshold common to a plurality of targets for determination, and a second determination to determine whether a target person is impersonating someone determined to be potentially impersonating someone in the first determination, based on a determination threshold adjusted for each of the targets for determination. 16. The information processing apparatus according to 15, characterized in that, in the first determination, in response to the detection of a target person who is potentially impersonating someone, the acquisition means acquires the status information of the target person for re-determination, the determination means determines the determination threshold of the target person for re-determination based on the status information, and the determination means performs the second determination on the target person for re-determination based on the determination threshold. 17. The information processing apparatus according to 15 or 16, characterized in that the acquisition means does not acquire the status information of the person to be judged who is determined not to be impersonating in the first judgment, the determination means does not determine the judgment threshold of the person to be judged, and the determination means does not perform the second judgment on the person to be judged. 18. The information processing apparatus according to any one of 1 to 17, characterized in that the determination means calculates an impersonation possibility score for each person to be judged, indicating the possibility that impersonation is occurring, and performs an impersonation judgment on the person to be judged based on the comparison result of the impersonation possibility score and the judgment threshold. 19. An information processing method comprising: one or more computers acquiring status information indicating at least one of surrounding circumstances, which are the circumstances of people around the person to be judged, and the person's own circumstances, which are the circumstances of the person to be judged, determining a judgment threshold for the impersonation judgment of the person to be judged based on the status information, and performing an impersonation judgment on the person to be judged based on the judgment threshold.20. A recording medium that records a program causing a computer to execute: an acquisition step of acquiring situational information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged; a determination step of determining a determination threshold for the impersonation determination of the person to be judged based on the situational information; and a determination step of performing the impersonation determination of the person to be judged based on the determination threshold.

[0227] Some or all of the appendices 2 to 18, which are dependent on the information processing device described in appendice 1 above, may also be dependent on the information processing method in appendice 19 and the recording medium in appendice 20 in the same dependent relationship as between appendice 1 and appendices 2 to 18. Furthermore, within the scope of not departing from each of the embodiments described above, some or all of the configurations described as appendices can be realized in various hardware, software, various recording means for recording software, or systems.

[0228] 10 Information processing device 10 11 Acquisition unit 12 Determination unit 13 Judgment unit 1A Processor 2A Memory 3A Input / Output I / F 4A Peripheral circuit 5A Bus

Claims

1. An information processing device comprising: an acquisition means for acquiring situation information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged; a determination means for determining a determination threshold for the impersonation determination of the person to be judged based on the situation information; and a determination means for performing the impersonation determination of the person to be judged based on the determination threshold.

2. The information processing apparatus according to claim 1, characterized in that the acquisition means acquires situational information indicating at least one of the following: whether or not there are people around the person to be determined, the number of people around the person to be determined, the distance between the people around the person to be determined and the person to be determined, the attributes of the people around the person to be determined, the posture of the people around the person to be determined, the direction of the faces of the people around the person to be determined, the gaze of the people around the person to be determined and the actions of the people around the person to be determined.

3. The information processing apparatus according to claim 1 or 2, characterized in that the acquisition means acquires situational information indicating at least one of the following: the posture of the person to be judged, the direction of the face of the person to be judged, the gaze of the person to be judged, and the actions of the person to be judged.

4. The information processing apparatus according to any one of claims 1 to 3, characterized in that the determination means determines the determination threshold based on the degree of attention focused on the person to be determined by surrounding persons who are people in the vicinity of the person to be determined.

5. The information processing apparatus according to claim 4, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not there is a surrounding person who has turned their gaze, face or body towards the person to be determined; the number of surrounding people who have turned their gaze, face or body towards the person to be determined; the number of surrounding people who have turned their gaze, face or body towards the person to be determined at the same time; a statistical value of the duration for which each of the multiple surrounding people has continued to turn their gaze, face or body towards the person to be determined; a statistical value of the cumulative time for which each of the multiple surrounding people has turned their gaze, face or body towards the person to be determined; a statistical value of the number of times each of the multiple surrounding people has turned their gaze, face or body towards the person to be determined; and the cumulative time for which at least one of the surrounding people has turned their gaze, face or body towards the person to be determined.

6. The information processing apparatus according to claim 4 or 5, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not there are any surrounding persons who have photographed the person to be determined; the number of surrounding persons who have photographed the person to be determined; the number of surrounding persons who have photographed the person to be determined simultaneously; a statistical value of the duration for which each of the multiple surrounding persons has continued to photograph the person to be determined; a statistical value of the cumulative time for which each of the multiple surrounding persons has photographed the person to be determined; a statistical value of the number of times each of the multiple surrounding persons has photographed the person to be determined; and the cumulative time for which at least one of the surrounding persons has photographed the person to be determined.

7. The information processing apparatus according to any one of claims 1 to 6, wherein the determination means determines the determination threshold based on at least one of the following: whether or not there are people with predetermined attributes around the person to be determined; the number of people with predetermined attributes around the person to be determined; the distance between the people with predetermined attributes around the person to be determined and the person to be determined; and statistical values ​​of the distance between each of the multiple people with predetermined attributes around the person to be determined and the person to be determined.

8. The information processing apparatus according to any one of claims 1 to 7, wherein the determination means determines the determination threshold based on at least one of the following: whether or not persons with predetermined attributes in the vicinity of the person to be determined are directing their gaze, face, or body towards the person to be determined; the number of persons with predetermined attributes who are directing their gaze, face, or body towards the person to be determined; the distance between the persons with predetermined attributes who are directing their gaze, face, or body towards the person to be determined and the person to be determined; and statistical values ​​of the distance between each of the multiple persons with predetermined attributes who are directing their gaze, face, or body towards the person to be determined and the person to be determined.

9. The information processing device according to claim 7 or 8, characterized in that the person having the predetermined attributes is a person who monitors the person to be determined.

10. The information processing apparatus according to any one of claims 1 to 9, characterized in that the determination means determines the determination threshold based on whether or not the person to be judged has performed a predefined suspicious behavior.

11. The information processing apparatus according to any one of claims 1 to 10, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not the person to be determined directed their gaze, face or body towards a predetermined person; the duration for which the person to be determined continued to direct their gaze, face or body towards the predetermined person; the cumulative duration for which the person to be determined directed their gaze, face or body towards the predetermined person; and the number of times the person to be determined directed their gaze, face or body towards the predetermined person.

12. The information processing apparatus according to any one of claims 1 to 10, characterized in that the determination means determines the determination threshold based on at least one of the following: whether or not the person to be determined did not turn their face and body towards a predetermined person but directed their gaze towards the predetermined person; the duration for which the person to be determined did not turn their face and body towards the predetermined person but continued to direct their gaze towards the predetermined person; the cumulative duration for which the person to be determined did not turn their face and body towards the predetermined person but directed their gaze towards the predetermined person; and the number of times the person to be determined did not turn their face and body towards the predetermined person but directed their gaze towards the predetermined person.

13. The information processing device according to claim 11 or 12, characterized in that the predetermined person is a person who monitors the person subject to determination.

14. The information processing apparatus according to any one of claims 1 to 13, characterized in that the determination means determines the determination threshold adjusted for each of the multiple persons to be determined, based on the status information of each of the persons to be determined.

15. The information processing apparatus according to any one of claims 1 to 14, characterized in that the determination means performs: a first determination to determine whether a target person is impersonating someone based on a predetermined common threshold common to a plurality of targets; and a second determination to determine whether a target person is impersonating someone who has been determined to be potentially impersonating someone in the first determination, based on a determination threshold adjusted for each of the targets.

16. The information processing apparatus according to claim 15, characterized in that, in the first determination, in response to the detection of a person subject to re-determination who is a person subject to determination who may be impersonating someone else, the acquisition means acquires the status information of the person subject to re-determination; the determination means determines the determination threshold of the person subject to re-determination based on the status information; and the determination means performs the second determination on the person subject to re-determination based on the determination threshold.

17. The information processing apparatus according to claim 15 or 16, characterized in that the acquisition means does not acquire status information of the person subject to the judgment who is determined to be the subject of the judgment who is determined to be the subject of the judgment who is determined to be the subject of the judgment who is determined to be the subject of the judgment who is determined to be the subject of the judgment who is determined to be the subject of the judgment, the determination means does not determine the judgment threshold of the person subject to the judgment who is determined to be the subject of the judgment who is determined to be the subject of the judgment, and the determination means does not perform the second judgment on the person subject to the judgment who is determined to be the subject of the judgment.

18. The information processing apparatus according to any one of claims 1 to 17, characterized in that the determination means calculates an impersonation possibility score for each of the targets to be determined, indicating the possibility that impersonation is occurring, and performs an impersonation determination of the target based on the result of comparing the impersonation possibility score with the determination threshold.

19. An information processing method comprising: one or more computers acquiring situational information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged; determining a judgment threshold for the impersonation judgment of the person to be judged based on the situational information; and performing the impersonation judgment of the person to be judged based on the judgment threshold.

20. A recording medium that records a program causing a computer to execute: an acquisition step of acquiring situational information indicating at least one of the surrounding situation, which is the situation of people around the person to be judged, and the person's own situation, which is the situation of the person to be judged; a determination step of determining a determination threshold for the impersonation judgment of the person to be judged based on the situational information; and a determination step of performing an impersonation judgment of the person to be judged based on the determination threshold.

Images