Information processing device, information processing method, and computer program

Abstract

An information processing device 1 comprises: an authentication unit 11 that performs authentication processing for determining success or failure of authentication of an object on the basis of a collation score generated by using biological information of the object and a prescribed condition; an inference unit 12 for inferring, on the basis of a past authentication result obtained by the authentication unit 11, inference authentication information on success or failure of authentication of the object when the authentication unit 11 repeats the authentication processing a prescribed number of times; and an output unit 13 for outputting the inference authentication information.

Description

Information processing device, information processing method, and recording medium

[0001] The present disclosure relates to the technical fields of an information processing device, an information processing method, and a recording medium.

[0002] Patent Document 1 describes a face recognition device including a face image acquisition unit that acquires a face image of an authentication target, a matching unit that performs face recognition by calculating a similarity between face information based on the face image of the authentication target and reference face information for each registered user and comparing the similarity with a threshold, a prediction unit that predicts a change in the similarity based on a history of similarities when authentication is successful, and a threshold change unit that changes the threshold based on the prediction result. Patent Document 2 describes a face recognition device including a face image acquisition unit that acquires a face image of an authentication target, a matching unit that performs face recognition by calculating a similarity between face information based on the face image of the authentication target and reference face information for each registered user and comparing the similarity with a threshold, a prediction unit that predicts a change in the similarity based on a history of similarities when authentication is successful, and a threshold change unit that changes the threshold based on the prediction result. Patent Document 3 describes an evaluation device that includes a generation unit that generates, from a first population of biometric data collected from a plurality of people, a plurality of sets that have a smaller number of biometric data items than a first population and that are at least partially different from each other; a first estimation unit that estimates a probability density distribution of similarity between stranger pairs for each of the plurality of sets; a calculation unit that calculates a stranger acceptance rate for each of the plurality of sets based on the probability density distribution, and a second estimation unit that estimates an integrated stranger acceptance rate from statistics of the stranger acceptance rates for each of the plurality of sets calculated by the calculation unit. Patent Document 4 describes a person recognition device that includes a mode selection means for selecting an operation mode in which a determination is made as to whether a person photographed by a camera is a registered person who has been registered in advance, or a setting mode in which a threshold value to be used when making the determination is set; a person determination means for determining, when the operation mode is selected, as a registered person if the similarity between extracted facial feature information and stored facial feature information is equal to or greater than a threshold value; an output means for outputting the determination result of the person determination means; and, when the setting mode is selected, a threshold value setting means for setting a threshold value based on the similarity between facial feature information extracted by photographing a registered person with a camera and the stored facial feature information, and the similarity between facial feature information extracted by photographing a person other than a registered person with a camera and the stored facial feature information.Patent Document 5 describes a biometric authentication device including a biometric information input unit that acquires subject biometric information, which is biometric information of the subject, from the subject of authentication; a biometric registration information storage unit that pre-stores user biometric information, which is biometric information of users of multiple biometric authentication devices including the subject; a matching unit that matches the subject biometric information acquired by the biometric information input unit with the user biometric information stored in the biometric information registration storage unit; an authentication history information storage unit that stores the matching result between the subject biometric information and the user biometric information as authentication history information; an authentication threshold determination unit that determines an authentication judgment threshold for authenticating the user by determining the degree of match of the matching result based on the authentication history information; and an authentication judgment unit that determines authentication of the subject based on the authentication judgment threshold determined by the authentication threshold determination unit and the biometric information acquired by the biometric information input unit. Patent Document 6 describes a technology in which wrinkle ridge information for the same finger of the same person to be authenticated is read m times, any two of these are selected from the mC2 combinations to calculate an error for each, similarity is calculated from these calculated errors, a characteristic curve for the person's FRR (false rejection rate) and a characteristic curve for the FAR (false acceptance rate) are found from this similarity, a threshold value specific to the person is set based on these characteristic curves, the wrinkle ridge information for the finger of the person to be authenticated is compared with pre-registered wrinkle ridge information for that person, and the set threshold value specific to the person is used to perform personal authentication.

[0003] International Publication No. 2019 / 138840 Japanese Patent Application Laid-Open No. 2019-125002 Japanese Patent Application Laid-Open No. 2018-028815 Japanese Patent Application Laid-Open No. 2019-125002 Japanese Patent Application Laid-Open No. 2012-123526 Japanese Patent Application Laid-Open No. 2001-21309

[0004] An object of this disclosure is to provide an information processing device, an information processing method, and a recording medium that aim to improve upon the techniques described in prior art documents.

[0005] One aspect of the information processing device includes an authentication means for performing an authentication process that determines whether the target will be authenticated based on a matching score generated using the target's biometric information and predetermined conditions, an estimation means for estimating estimated authentication information regarding whether the target will be authenticated when the authentication means repeats the authentication process a predetermined number of times based on past authentication results by the authentication means, and an output means for outputting the estimated authentication information.

[0006] In one aspect of the information processing method, an authentication means performs an authentication process to determine whether the target will be authenticated based on a matching score generated using the target's biometric information and predetermined conditions, and based on past authentication results by the authentication means, estimates estimated authentication information regarding whether the target will be authenticated when the authentication means repeats the authentication process up to a predetermined number of times, and outputs the estimated authentication information.

[0007] In one aspect of the recording medium, a computer program is recorded on a computer to cause a computer to execute an information processing method in which an authentication means performs an authentication process to determine whether or not the authentication of the subject will be successful based on a matching score generated using the subject's biometric information and predetermined conditions, estimates estimated authentication information regarding whether or not the authentication of the subject will be successful when the authentication means repeats the authentication process up to a predetermined number of times based on past authentication results by the authentication means, and outputs the estimated authentication information.

[0008] FIG. 1 is a block diagram showing the configuration of an information processing device in a first embodiment. FIG. 2 is a block diagram showing the configuration of an information processing device in a second embodiment. FIG. 3 is a flowchart showing the flow of information processing operation of the information processing device in the second embodiment. FIG. 4 is a conceptual diagram of the information processing operation of the information processing device in the second embodiment. FIG. 5 is a flowchart showing the flow of information processing operation of the information processing device in the second embodiment. FIG. 6 is a flowchart showing the flow of information processing operation of the information processing device in the second embodiment. FIG. 7 is a block diagram showing the configuration of an information processing device in a third embodiment. FIG. 8 is a flowchart showing the flow of information processing operation of the information processing device in the third embodiment. FIG. 9 is a flowchart showing the flow of information processing operation of the information processing device in the third embodiment. FIG. 10 is a flowchart showing the flow of information processing operation of the information processing device in the third embodiment. FIG. 11 is a block diagram showing the configuration of an information processing device in a fourth embodiment. FIG. 12 is a flowchart showing the flow of information processing operation of the information processing device in the fourth embodiment. FIG. 13 is a conceptual diagram showing an example of output of the information processing device in the fourth embodiment. FIG. 14 is a block diagram showing the configuration of an information processing device in a fifth embodiment. FIG. 15 is a flowchart showing the flow of information processing operation of the information processing device in the fifth embodiment. Fig. 16 is a conceptual diagram showing an example of output from an information processing device in the fifth embodiment. Fig. 17 is a block diagram showing the configuration of an information processing device in the sixth embodiment. Fig. 18 is a conceptual diagram showing an example of output from an information processing device in the sixth embodiment. Fig. 19 is a block diagram showing the configuration of an information processing device in the seventh embodiment.

[0009] Hereinafter, embodiments of an information processing device, an information processing method, and a recording medium will be described with reference to the drawings. [1: First Embodiment]

[0010] A first embodiment of an information processing device, an information processing method, and a recording medium will be described below. Hereinafter, the first embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 1 to which the first embodiment of the information processing device, the information processing method, and the recording medium is applied. [1-1: Configuration of Information Processing Device 1]

[0011] FIG. 1 is a block diagram showing the configuration of an information processing device 1 in the first embodiment. As shown in FIG. 1, the information processing device 1 includes an authentication unit 11, an estimation unit 12, and an output unit 13. The authentication unit 11 performs authentication processing to determine whether authentication of the target is successful or not based on a matching score generated using biometric information of the target and predetermined conditions. The estimation unit 12 estimates estimated authentication information related to whether authentication of the target is successful or not when the authentication unit 11 repeats the authentication processing up to a predetermined number of times based on past authentication results by the authentication unit 11. The output unit 13 outputs the estimated authentication information. [1-2: Technical Effects of the Information Processing Device 1]

[0012] The information processing device 1 according to the first embodiment can provide information on whether or not the target authentication will be successful when the authentication process is repeated a predetermined number of times, estimated based on past authentication results. [2: Second Embodiment]

[0013] Next, a second embodiment of the information processing device, the information processing method, and the recording medium will be described. Hereinafter, the second embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 2 to which the second embodiment of the information processing device, the information processing method, and the recording medium is applied. [2-1: Configuration of the information processing device 2]

[0014] 2 is a block diagram showing the configuration of an information processing device 2 in the second embodiment. As shown in FIG. 2, the information processing device 2 includes a calculation device 21 and a storage device 22. The information processing device 2 may further include a communication device 23, an input device 24, and an output device 25. However, the information processing device 2 does not necessarily have to include at least one of the communication device 23, the input device 24, and the output device 25. The calculation device 21, the storage device 22, the communication device 23, the input device 24, and the output device 25 may be connected via a data bus 26.

[0015] The arithmetic device 21 includes, for example, at least one of a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), and an FPGA (Field Programmable Gate Array). The arithmetic device 21 reads a computer program. For example, the arithmetic device 21 may read a computer program stored in the storage device 22. For example, the arithmetic device 21 may read a computer program stored in a computer-readable, non-transitory recording medium using a recording medium reading device (e.g., an input device 24 described later) not shown in the drawings that is included in the information processing device 2. The arithmetic device 21 may acquire (i.e., download or read) the computer program from a device (not shown) located outside the information processing device 2 via the communication device 23 (or another communication device). The arithmetic device 21 executes the read computer program. As a result, logical functional blocks for executing the operations to be performed by the information processing device 2 are realized within the arithmetic device 21. In other words, the arithmetic device 21 can function as a controller for realizing logical functional blocks for executing the operations (in other words, processing) to be performed by the information processing device 2.

[0016] 2 shows an example of logical functional blocks implemented within the computing device 21 to perform information processing operations. As shown in FIG. 2 , the computing device 21 implements an authentication unit 211, which is a specific example of "authentication means" described in the appendix, an estimation unit 212, which is a specific example of "estimation means" described in the appendix, and an output control unit 213, which is a specific example of "output means" described in the appendix. The authentication unit 211 may include an acquisition unit 2111, a matching score generation unit 2112, and a success / failure determination unit 2113. The estimation unit 212 may include an authentication success / failure probability calculation unit 2121. Details of the operations of the authentication unit 211, the estimation unit 212, and the output control unit 213 will be described later with reference to FIG. 4.

[0017] The storage device 22 can store desired data. For example, the storage device 22 may temporarily store a computer program executed by the arithmetic device 21. The storage device 22 may temporarily store data that the arithmetic device 21 temporarily uses when the arithmetic device 21 is executing a computer program. The storage device 22 may store data that the information processing device 2 stores for a long period of time. The storage device 22 may include at least one of a RAM (Random Access Memory), a ROM (Read Only Memory), a hard disk device, a magneto-optical disk device, an SSD (Solid State Drive), and a disk array device. In other words, the storage device 22 may include a non-transitory recording medium. A biometric information registration unit 221 and an authentication result storage unit 222 may be implemented within the storage device 22.

[0018] The communication device 23 is capable of communicating with devices external to the information processing device 2 via a communication network (not shown). The communication device 23 may be a communication interface based on standards such as Ethernet (registered trademark), Wi-Fi (registered trademark), Bluetooth (registered trademark), or USB (Universal Serial Bus).

[0019] The input device 24 is a device that accepts information input to the information processing device 2 from outside the information processing device 2. For example, the input device 24 may include an operation device (e.g., at least one of a keyboard, a mouse, and a touch panel) that can be operated by an operator of the information processing device 2. For example, the input device 24 may include a reading device that can read information recorded as data on a recording medium that can be externally attached to the information processing device 2.

[0020] The output device 25 is a device that outputs information to the outside of the information processing device 2. For example, the output device 25 may output information as an image. That is, the output device 25 may include a display device (a so-called display) that can display an image showing the information to be output. For example, the output device 25 may output information as sound. That is, the output device 25 may include an audio device (a so-called speaker) that can output sound. For example, the output device 25 may output information on paper. That is, the output device 25 may include a printing device (a so-called printer) that can print desired information on paper. [2-2: Authentication Operation Performed by Information Processing Device 2]

[0021] The flow of the authentication operation performed by the information processing device 2 will be described with reference to Fig. 3. Fig. 3 is a flowchart showing the flow of the authentication operation performed by the information processing device 2.

[0022] As shown in FIG. 3, the authentication unit 211 sets "M" to "0" (step S10). "M" indicates the number of times the authentication process is repeated. The repetition of the authentication process ends when the matching score exceeds the matching score threshold. Alternatively, the repetition of the authentication process ends when the upper limit number of repetitions "M ｍａｘ The authentication process may be repeated at predetermined time intervals. Alternatively, the authentication process may be repeated for each distance between the target and the camera capturing the target.

[0023] The authentication unit 211 increments "M" by "1" (step S11). ｍａｘ It is determined whether or not the value exceeds the predetermined value (step S12).

[0024] "M" is "M ｍａｘ " (Step S12: No), the authentication unit 211 performs authentication processing. The authentication processing is processing for determining whether the authentication of the target is successful or not based on the matching score and predetermined conditions. The predetermined conditions include a condition for determining that the authentication of the target is successful and terminating the repetition of the authentication processing when the matching score exceeds the threshold value of the matching score, and a condition for determining that the authentication of the target is successful and terminating the repetition of the authentication processing when the matching score exceeds the threshold value of the matching score. ｍａｘ " Repeatedly, it contains a condition to end the repeat.

[0025] The acquisition unit 2111 acquires biometric information of the target (step S13). The acquisition unit 2111 may acquire the biometric information of the target photographed by a camera. In this case, the biometric information may be a biometric image such as a face image, an iris image, a fingerprint image, or a vein image. The biometric information may also be a feature extracted from the biometric image. The acquisition unit 2111 may acquire the biometric information of the target collected by a microphone. In this case, the biometric information may be audio information or the like.

[0026] The matching score generation unit 2112 generates a matching score using the biometric information (step S14). The matching score generation unit 2112 generates a matching score based on the acquired biometric information and the biometric information registered in the biometric information registration unit 221. The matching score is a score according to the result of matching the acquired biometric information with the registered biometric information. The matching score represents the degree of matching. The matching score may represent the degree of agreement between the biometric information. The matching score may also represent the degree of similarity between the biometric information. In this case, a similarity score may be generated as the matching score.

[0027] The success / failure determination unit 2113 determines whether the matching score exceeds the matching score threshold (step S15). If the matching score exceeds the threshold (step S15: Yes), the authentication unit 211 determines that the authentication of the target is successful (step S16).

[0028] If the matching score does not exceed the threshold (step S15: No), the process returns to step S11 and the above authentication process is performed. ｍａｘ " (step S12: Yes), the authentication unit 211 determines that the target authentication has failed (step S17).

[0029] The authentication unit 211 stores the authentication result in the authentication result storage unit 222 (step S18). The past authentication results stored in the authentication unit 211 include information indicating the end of the iteration of the authentication process, and the matching score at the end of the iteration (referred to as the "end matching score").

[0030] In addition, the maximum number of repetitions is "M ｍａｘ" may be set according to, for example, the characteristics of the authentication mechanism used by the authentication unit 211, the environment in which the authentication is performed, the purpose of the authentication, etc. The upper limit number of repetitions may be set to, for example, a number at which no increase in the matching score is expected even if the authentication process is repeated any further.

[0031] FIG. 4( a) illustrates an example of an authentication result when an object is authenticated while moving toward the camera. The horizontal axis illustrated in FIG. 4( a) may represent, for example, a quantity determined based on the distance between the object approaching the camera and the camera (however, the quantity decreases as the distance increases, such as the inverse of the distance). The horizontal axis illustrated in FIG. 4( a) may represent a quantity representing the size of the object in the image, such as the distance between the object's eyes or the size of the object's face in the image. Alternatively, the horizontal axis illustrated in FIG. 4( a) may represent time. The vertical axis illustrated in FIG. 4( a) represents the magnitude of the matching score, which indicates the degree of matching. Also, the magnitude of the matching score threshold TH0 used when the corresponding authentication process was performed is illustrated parallel to the horizontal axis.

[0032] 4(a) illustrates a case where the matching score generated in the first authentication process and the matching score generated in the second authentication process do not exceed a set threshold TH0, and the matching score generated in the third authentication process exceeds the threshold TH0, resulting in the repetition being terminated. In the example illustrated in FIG. 4(a), the matching score generated in the third authentication process exceeds the threshold TH0, so the end of the repetition is three. In the example illustrated in FIG. 4(a), the fourth and subsequent authentication processes have not been performed, and the matching score when the fourth and subsequent authentication processes are performed is unknown.

[0033] In this embodiment, the estimation unit 212 estimates estimated authentication information regarding the success or failure of target authentication when the authentication unit 211 repeats the authentication process a predetermined number of times, based on past authentication results by the authentication unit 211. The predetermined number of times may be a number greater than the end number of times. The predetermined number of times may also be the above-mentioned upper limit number of repetitions.

[0034] 4(b), if the success or failure of matching is determined using a threshold value TH1 greater than TH0, the matching score in the third authentication process does not exceed the threshold value TH1, and therefore the fourth or subsequent authentication processes are performed. The information processing device 2 in this embodiment calculates the probability of occurrence of a matching score generated in an authentication process that was not actually performed.

[0035] In this embodiment, the occurrence probability of the next matching score generated in the authentication process performed after the final round may be information corresponding to the final matching score of the final round. As illustrated in FIG. 4B, in a case where the matching score increases with each execution of the authentication process, the next matching score generated in the authentication process performed after the final round may have a high probability of being larger than the final matching score of the final round.

[0036] As illustrated in FIG. 4( b), the estimation unit 212 may calculate the probability h of occurrence of the matching score s in the fourth authentication process. For example, FIG. 4( b) illustrates that the probability h of occurrence of the matching score s indicated by the second arrow from the top is greater than the probability h of occurrence of the matching score s indicated by the top arrow. FIG. 4( b) also illustrates that the probability h of occurrence of the matching score s indicated by the third arrow from the top is smaller than the probability h of occurrence of the matching score s indicated by the second arrow from the top. FIG. 4( b) also illustrates that the probability h of occurrence of the matching score s indicated by the bottom arrow is smaller than the probability h of occurrence of the matching score s indicated by the third arrow from the top.

[0037] The occurrence probability h may change depending on the number of times the matching process is performed. In other words, the occurrence probability of the matching score obtained in the fifth matching process may be different from the occurrence probability of the matching score obtained in the fourth matching process. The estimation unit 212 uses the occurrence probability of the next matching score to determine the occurrence probability of the matching score that will be generated when the authentication process is performed after the final round. By repeating this process, the estimation unit 212 can determine, depending on the final matching score, the occurrence probability of the matching score that will be generated when the authentication process is repeated a predetermined number of times after the final round.

[0038] The estimation unit 212 may calculate the occurrence probability using an occurrence probability estimation model, which will be described in another embodiment.

[0039] 4(b), when the threshold is raised to TH1, the matching failure probability p2 at which the matching score generated by the fourth authentication process will be equal to or less than TH1 can be calculated by integrating the occurrence probability h of the matching score s from 0 to TH1. Similarly, the matching success probability p1 at which the matching score generated by the fourth authentication process will exceed TH1 can be calculated by integrating the occurrence probability h when the matching score s exceeds TH1. In actual calculations, the integral value can be calculated by the quadrature by division method, which divides the occurrence probability distribution into small sections and calculates and adds up the areas.

[0040] [2-3: Identity Verification Score / Other Identity Verification Score]

[0041] A matching score generated using biometric information of the same subject is called an "identity matching score." In one-to-N authentication, registered biometric information includes biometric information of the subject corresponding to the captured biometric information and biometric information of other subjects. Among the matching scores generated using the captured biometric information and the registered biometric information, the matching score generated using the biometric information of the same subject is estimated to be the highest. However, this does not necessarily hold true in cases where a person is mistakenly matched with another person. When other biometric information can be used to determine whether a person is the real person, such as in other authentication methods or multimodal authentication, the score obtained when a person is determined to be the real person is called an "identity matching score." Specifically, during the initial adjustment period after system implementation, it is possible to determine whether a person is the real person using other authentication methods that are not used during normal operation. Alternatively, when there are multiple authentication methods (such as when passing through multiple gates or capturing a person multiple times with multiple cameras) (each method may use different authentication methods), the matching results from multiple gates can be integrated to determine whether a person is the real person. On the other hand, if no other authentication means are available, the largest matching score among the matching scores generated using the captured biometric information and the registered biometric information may be considered as the "identity matching score."

[0042] Furthermore, a matching score generated using biometric information of different subjects is called an "other person matching score." Of the matching scores generated using the captured biometric information and the registered biometric information, it can be assumed that the matching score generated using the biometric information of different subjects will not be the largest, but will be the second largest or smaller. However, this also does not hold true in cases where a different person is mistakenly matched. For this reason, the score obtained when the person is determined to be other than the person in question, as described above, is called an "other person matching score." On the other hand, if no other authentication means are available, the second largest or smaller matching score among the matching scores generated using the captured biometric information and the registered biometric information may be considered an "other person matching score."

[0043] The authentication unit 211 may perform one-to-N authentication. That is, the authentication unit 211 may perform authentication processing between N registered biometric information pieces including the biometric information of the person in question and the acquired biometric information. The matching score generation unit 2112 may generate N matching scores in one authentication processing. Hereinafter, among the matching scores generated by the matching score generation unit 2112, the matching score determined to be the person in question (usually the largest matching score) is treated as the person in question matching score. Furthermore, among the matching scores generated by the matching score generation unit 2112, the largest score determined to be a different person (usually the second largest matching score) is treated as the different person matching score. [2-4: Probability of failure of person in person matching performed by the information processing device 2 "P ｆ " Calculation operation]

[0044] The flow of the operation performed by the information processing device 2 to calculate at least one of the probability of failure in identity verification and the probability of success in identity verification (the probability of falsely identifying a different person as successful in authentication) as estimated authentication information will be described with reference to Fig. 5 and Fig. 6. Fig. 5(a) is a flowchart showing the flow of the operation performed by the information processing device 2 to calculate the probability of failure in identity verification, and Fig. 5(b) is a flowchart showing the flow of the operation performed by the information processing device 2 to calculate the probability of success in identity verification. Fig. 6 is a flowchart showing the flow of the operation performed by the information processing device 2 to calculate the probability of failure in identity verification and the probability of success in identity verification.

[0045] In the flow shown in FIG. 5(a), the personal identification score is handled, and the personal identification failure probability "P ｆ Calculate the probability of failure in personal identification "P ｆ " indicates the probability of failing to match the person.

[0046] 5A, the authentication success / failure probability calculation unit 2121 acquires the authentication result stored in the authentication result storage unit 222 (step S20). The authentication result is stored in the information "M ０ ", and the final identity verification score "s Ｍ０ " It includes "M ０ The distribution of the identity verification scores generated in the "th authentication process" may be expressed by a delta function as shown in the following formula 1. [Formula 1]

[0047] The authentication success / failure probability calculation unit 2121 converts "M" into "M ０ +1" (step S21). "M" indicates the number of times the authentication process is repeated. ０ " indicates the end of the authentication process repetition, so "M ０ +1" indicates the next number of times for the final round. For example, in the case of the example shown in FIG. ０ " is "3" and "M ０ "+1" is "4".

[0048] The authentication success / failure probability calculation unit 2121 calculates whether "M" is "M ｍａｘ It is determined whether or not "M ｍａｘ " indicates the upper limit of the number of repetitions.

[0049] "M" is "M ｍａｘ " (Step S22: No), the authentication success / failure probability calculation unit 2121 calculates the distribution of the identity verification score in the Mth authentication process (Step S23).

[0050] Final episode "M ０ If the personal verification score in "M ０ The conditional occurrence probability of the identity verification score of "+1" is expressed as "h(s|x)". ０+1” identity verification score distribution “P ｓｅｌｆ (M ０ +1, s) may be expressed by the following formula 2. Ｍ０ is M ０ is the identity verification score for the th round. [Equation 2]

[0051] Therefore, the next "M ０ +1) and later, the distribution of identity verification scores "P ｓｅｌｆ (M, s)) may be expressed by the following formula 3. That is, the next "M ０ +1) and afterwards, the distribution of identity verification scores "P ｓｅｌｆ (M, s) may be calculated from the occurrence probability distribution of the matching score in which the previous (M-1)th matching score in the "M"th time is equal to or less than the threshold value TH, and the conditional occurrence probability distribution of the matching score in the next time for each matching score. In step S23, the distribution of the matching score "P ｓｅｌｆ (M, s)" may be calculated. [Equation 3] The conditional occurrence probability distribution h(s|x) used here may be calculated and stored in advance as an occurrence probability estimation model for the identity verification score, as will be described later. Note that although the occurrence probability distribution here depends only on the identity verification score of the previous time, a model that is determined depending also on the identity verification scores prior to that may be used. For example, instead of the identity verification score of the previous time, the distribution may depend on a statistical representative value (such as the average or median) of the identity verification scores prior to that.

[0052] The authentication success / failure probability calculation unit 2121 increments "M" by "1" (step S24). Every time the authentication success / failure probability calculation unit 2121 increments "M" by "1", it calculates the distribution of the identity verification score expressed by the above formula 3.

[0053] "M" is "M ｍａｘ " (Step S22: Yes), the authentication success / failure probability calculation unit 2121 calculates the probability of failure of identity verification (Step S25). ｍａｘ Distribution of identity verification scores in the case of "P ｓｅｌｆ (Mｍａｘ , s)) The authentication success / failure probability calculation unit 2121 may calculate the authentication success / failure probability expressed by the above formula 3 as ｍａｘ -M ０ The probability of failure in personal identification can be calculated by repeating the process "M times." ｍａｘ ” is the probability that the personal verification score does not exceed the threshold TH and authentication fails. ｆ " may be expressed by the following formula 4. [Formula 4]

[0054] That is, the authentication success / failure probability calculation unit 2121 uses the occurrence probability to calculate the identity matching failure probability "P ｆ ” can be estimated. [2-5: The probability of success in matching another person performed by the information processing device 2 “P ｓ " Calculation operation]

[0055] In the flow shown in FIG. 5(b), the other person matching score is handled, and the other person matching success probability "P ｓ ” is calculated. ｓ " indicates the probability of successfully matching with someone other than the person.

[0056] As shown in Fig. 5B, the authentication success / failure probability calculation unit 2121 acquires the authentication result stored in the authentication result storage unit 222 (step S20). The authentication result is stored in the information "M ０ ", and the final other matching score "t Ｍ０ " It includes "M ０ The distribution of the other person matching scores generated in the "th" authentication process may be expressed by a delta function as shown in the following formula 5. [Formula 5]

[0057] The authentication success / failure probability calculation unit 2121 converts "M" into "M ０ +1" (step S21). ０ " indicates the end of the authentication process repetition, so "M ０"+1" indicates the next number of times after the end of the round.

[0058] "M" is "M ｍａｘ " (Step S22: No), the authentication success / failure probability calculation unit 2121 calculates the distribution of the other person matching score in the Mth authentication process (Step S26).

[0059] When the occurrence probability of the other person matching score is expressed as "g(t)", the next "M ０ +1” ｏｔｈｅｒ (M ０ +1, t) may be expressed by the following formula 6. In this embodiment, it is assumed that the occurrence probability g of the other person matching score does not depend on the previous other person matching score "y" (g(t|y) = g(t)). [Formula 6] Therefore, the next "M ０ The distribution of the match score after "+1" ｏｔｈｅｒ (M, t)) may be expressed by the following formula 7. ０ Distribution of other matching scores after "P ｏｔｈｅｒ (M, t) may be calculated from the probability that the other person matching score of the previous (M-1)th time is equal to or less than the threshold value TH and the occurrence probability distribution of the matching score of the next time. In step S26, the distribution of other person matching scores "P ｏｔｈｅｒ (M, t)" may be calculated. [Equation 7] The occurrence probability distribution g(t) used here may be calculated in advance and stored as an occurrence probability estimation model for the other person matching score, as will be described later. Note that although the occurrence probability distribution is assumed not to depend on the other person matching score from the previous time, it may be determined depending on the other person matching score from the previous time, as in the case of the identity matching score. In that case, an occurrence probability estimation model similar to the identity matching score may be generated, stored, and used. Furthermore, a model determined depending on the other person matching score from earlier times may also be used.

[0060] The authentication success / failure probability calculation unit 2121 increments "M" by "1" (step S24). Every time the authentication success / failure probability calculation unit 2121 increments "M" by "1", it calculates the distribution of the other person matching score expressed by the above formula 7.

[0061] "M" is "M ｍａｘ " (step S22: Yes), the authentication success / failure probability calculation unit 2121 calculates the probability of successful matching against another person (step S27). ０ +1" to "M ｍａｘ Distribution of other matching scores up to "P ｏｔｈｅｒ The authentication success / failure probability calculation unit 2121 may calculate the probability of successful other person matching by adding up the probabilities that the other person matching score is equal to or greater than the threshold value TH in each iteration. The authentication success / failure probability calculation unit 2121 may calculate the probability of successful other person matching by performing the process expressed by the above formula 7 using "M ｍａｘ -M ０ The probability of successful matching of another person can be calculated by repeating the process "M times." ｍａｘ The probability that the false matching score exceeds the threshold TH and false acceptance occurs before reaching the false matching success probability "P ｓ " may be expressed by the following formula 8. [Formula 8] That is, the authentication success / failure probability calculation unit 2121 uses the occurrence probability to calculate the other person matching success probability "P ｓ 2-6: Operation of Calculating the Probability of Failure in Identity Verification and the Probability of Success in Identity Verification Performed by the Information Processing Device 2

[0062] Next, a calculation operation will be described when both rejection of the actual person and acceptance of other people are taken into consideration during authentication processing. In the flow shown in Fig. 6(a), the actual person matching score and the other person matching score are handled, and the probability of failure of actual person matching and the probability of success of other person matching when the upper limit of repetitions is reached are calculated.

[0063] As shown in FIG. 6A, the authentication success / failure probability calculation unit 2121 acquires the authentication result stored in the authentication result storage unit 222 (step S20). ０ The distribution of the true person matching score and the distribution of the false person matching score generated in the "th authentication process" may be formulated as in the following Equation 9. [Equation 9]

[0064] The authentication success / failure probability calculation unit 2121 converts "M" into "M ０ +1" (step S21). ｍａｘ " (Step S22: No), the authentication success / failure probability calculation unit 2121 calculates the distribution of the true person matching score and the distribution of the false person matching score in the Mth authentication process (Step S28).

[0065] The next episode of the final episode, "M ０ The distribution of the true identity matching score and the distribution of the false identity matching score of "+1" may be formulated as in the following Equation 10. [Equation 10]

[0066] The next episode of "M ０ The distribution of the true person matching score and the distribution of the false person matching score from the "+1" time onwards may be formulated as in the following Equation 11. [Equation 11]

[0067] The authentication success / failure probability calculation unit 2121 increments "M" by "1" (step S24). Every time "M" is incremented by "1", the authentication success / failure probability calculation unit 2121 calculates the distribution of the true person matching score and the distribution of the false person matching score expressed by the above formula 11. ｍａｘ " (Step S22: Yes), the authentication success / failure probability calculation unit 2121 calculates the probability of failure in identity matching and the probability of success in false matching (Step S29).

[0068] The probability of failure in identity verification when considering both the person's rejection and the acceptance of others is "P ｆ " may be expressed as the following formula 12. [Formula 12]

[0069] That is, the probability of failure in personal identification "P ｆ" may represent the probability of a false negative, in which neither the true person matching score nor the false person matching score exceeds the threshold value TH, when the repetition is repeated up to the upper limit number of times.

[0070] In addition, the probability of successful matching when considering both the person's rejection and the acceptance of others is "P ｓ " may be expressed as in the following formula 13. [Formula 13]

[0071] Probability of successful matching with others "P ｓ " may represent the sum of the probability that the other person matching score exceeds the threshold and the identity matching score does not exceed the threshold, and the probability that the other person matching score exceeds the threshold and the identity matching score exceeds the threshold but is smaller than the other person matching score.

[0072] 6(b) is a conceptual diagram when both rejection by the actual person and acceptance by others are taken into consideration. The probability of failure in identity verification can be expressed by a region PF where both the identity verification score and the identity verification score do not exceed the threshold TH. The probability of success in identity verification can be expressed by a region PSa where the identity verification score exceeds the threshold but does not exceed the threshold, and a region PSb where the identity verification score exceeds the threshold but is smaller than the identity verification score.

[0073] It should be noted that even if it is possible to detect that the person in question has been rejected, it may be difficult to detect that a stranger has been accepted. In such cases, there may be cases where both the person in question and the stranger matching score can be used, or where only one of them can be used. When only one of them can be used, one of them is calculated as described with reference to FIG. 5.

[0074] The output control unit 213 calculates the probability of failure in identity verification "P" when the authentication process is repeated a predetermined number of times at a predetermined threshold value of the verification score. ｆ " and the probability of successful matching with another person "P ｓ4B, the output control unit 213 may change the threshold value from TH0 to TH1 and cause the display serving as the output device 25 to display visually easy-to-understand information about the possible values ​​of the matching score when further authentication processing is repeated and the probability of occurrence of the possible values.

[0075] Furthermore, the information processing device 2 determines the optimum upper limit number of repetitions "M ｍａｘ " may be asked for.

[0076] The authentication unit 211 does not have to include the acquisition unit 2111 and the matching score generation unit 2112. In other words, the authentication unit 211 may be a mechanism that performs authentication processing to determine whether the target authentication is successful or not based on the matching score and predetermined conditions. In this case, the authentication unit 211 acquires the matching score from a mechanism external to the authentication unit 211. The external mechanism is a mechanism equivalent to the acquisition unit 2111 and the matching score generation unit 2112. [2-7: Technical Effects of Information Processing Device 2]

[0077] The information processing device 2 in the second embodiment calculates the probability of occurrence of a matching score generated when authentication processing is performed after the final round according to the final matching score, and can therefore accurately estimate information relating to the success or failure of target authentication when the authentication processing is repeated a predetermined number of times. Because the information processing device 2 uses the occurrence probability, it can accurately estimate at least one of the probability of failure in true person matching and the probability of success in false person matching when the authentication processing is repeated a predetermined number of times at a predetermined threshold value for the matching score. Furthermore, the information processing device 2 ends the repetition of authentication processing when the matching score exceeds the threshold value for the matching score, and therefore can suppress unnecessary authentication processing. [3: Third Embodiment]

[0078] Next, a third embodiment of the information processing device, the information processing method, and the recording medium will be described. Hereinafter, the third embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 3 to which the third embodiment of the information processing device, the information processing method, and the recording medium is applied. [3-1: Configuration of the information processing device 3]

[0079] The configuration of the information processing device 3 in the third embodiment will be described with reference to Fig. 7. Fig. 7 is a block diagram showing the configuration of the information processing device 3 in the third embodiment.

[0080] 7 , the information processing device 3 in the third embodiment differs from the information processing device 2 in the second embodiment in that the estimation unit 312 includes a false authentication rate calculation unit 3122. Other features of the information processing device 3 may be the same as other features of the information processing device 2. Therefore, hereinafter, only the parts that differ from the embodiments already described will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate.

[0081] The false authentication rate calculation unit 3122 performs at least one of the following operations: an operation of estimating a false negative identification rate (FNIR, False Negative Identification-error Rate, hereinafter simply referred to as the false reject rate) at which the correct person is mistakenly rejected as a whole, using the probability of false matching failure for each authentication operation described in the second embodiment, at a predetermined threshold value of the matching score; and an operation of estimating a false positive identification rate (FPIR, False Positive Identification-error Rate, hereinafter simply referred to as the false acceptance rate) at which a wrong person is mistakenly accepted as a whole, using the probability of false matching success for each authentication operation described in the second embodiment. The false authentication rate calculation unit 3122 sets the predetermined threshold value to a desired value. The predetermined threshold value may be set to a value greater than the threshold value used in past authentication processes.

[0082] The authentication result storage unit 322 may store the results of K past authentication operations (a collection of a series of repeated authentications for the same subject up to the maximum number of repetitions). The K past authentication results may be the results of authentication for a total of K people. In other words, the K past authentication results may include authentication results for the same person. Furthermore, the K past authentication results may all include authentication results for different people.

[0083] Each authentication result includes information indicating the final round, and the true person verification score and the false person verification score in the final round, or the largest and second largest final matching scores among the final matching scores, as the true person verification score and the false person verification score, respectively. The true person verification score among the final matching scores in the final round is called the "final true person verification score." Also, the false person verification score in the final round is called the "final false person verification score." [3-2: Calculation operation of false rejection rate performed by information processing device 3]

[0084] In this embodiment, the number of times of false rejection is "S ｆ " is the probability of authentication failure P ｆ The number of times is calculated by adding up the number of times for K authentication operations. ｆ " is the number of times obtained by adding values ​​calculated probabilistically, so it is not necessarily an integer and is usually a decimal value.

[0085] As shown in FIG. 8, the false authentication rate calculation unit 3122 calculates "L" and the number of false rejections "S ｆ " to "0" (step S30). The false authentication rate calculation unit 3122 acquires past authentication results from the authentication result storage unit 322 (step S31). The estimation unit 312 increments "L" by "1" (step S32).

[0086] The false authentication rate calculation unit 3122 determines whether the final identity matching score included in the past authentication result exceeds a predetermined threshold (step S33). If the final identity matching score exceeds the predetermined threshold (step S33: Yes), it determines whether L is less than K (step S34). That is, it determines whether there are any unprocessed past authentication results among the K past authentication results. If there are any unprocessed past authentication results (step S34: Yes), the process returns to step S31 and performs processing using the next past authentication result.

[0087] If the final identity verification score does not exceed the predetermined threshold (step S33: No), the false authentication rate calculation unit 3122 ０ " is "M ｍａｘ" (Step S35). ０ " is "M ｍａｘ ” (step S35: No), the authentication success / failure probability calculation unit 2121 calculates the personal verification failure probability P ｆ (Step S36). ０ " is "M ｍａｘ " (step S35: Yes), the false authentication rate calculation unit 3122 calculates "P ｆ " is set to "1" (step S37).

[0088] The false authentication rate calculation unit 3122 calculates the number of false rejections "S ｆ " to "P ｆ " are added together (step S38). If there is an unprocessed past authentication result among the K past authentication results (step S34: Yes), the process returns to step S31 and the next past authentication result is used for processing. If there is no unprocessed past authentication result (step S34: No), the false authentication rate calculation unit 3122 calculates the false rejection rate of falsely rejecting the correct person at a predetermined threshold (step S39). The false authentication rate calculation unit 3122 calculates the false rejection rate of falsely rejecting the correct person at a predetermined threshold (step S39). ｆ Calculate the false rejection rate by dividing " by "K".

[0089] That is, the false authentication rate calculation unit 3122 can estimate the false rejection rate of falsely rejecting a person at a predetermined threshold value of the matching score using the matching failure probability. [3-3: Calculation of false acceptance rate performed by the information processing device 3]

[0090] In this embodiment, the number of times of accepting a third party is “O Ｓ " is the number of times that the final false matching score exceeds the predetermined threshold, and the number of times that the false matching score exceeds the predetermined threshold, when the final false matching score does not exceed the predetermined threshold, by adding up the probability that the false matching score will exceed the predetermined threshold over K authentication operations until the authentication process is repeated up to the upper limit number of times. Ｓ " is not necessarily an integer, but is usually a decimal value.

[0091] As shown in FIG. 9, the false authentication rate calculation unit 3122 calculates "L" and the number of false acceptances "O Ｓ" to "0" (step S30). The false authentication rate calculation unit 3122 acquires past authentication results from the authentication result storage unit 322 (step S31). The false authentication rate calculation unit 3122 increments "L" by "1" (step S32).

[0092] The false authentication rate calculation unit 3122 determines whether the final false authentication score included in the past authentication result exceeds a predetermined threshold (step S40). If the final false authentication score exceeds the predetermined threshold (step S40: Yes), the false authentication rate calculation unit 3122 determines whether the final false authentication score exceeds the predetermined threshold (step S40: Yes), Ｓ " is incremented by "1" (step S41).

[0093] If the final other person matching score does not exceed the predetermined threshold (step S40: No), the false authentication rate calculation unit 3122 ０ " is "M ｍａｘ " (Step S35). ０ " is "M ｍａｘ ” (step S35: No), the authentication success / failure probability calculation unit 2121 calculates the other person matching success probability P ｓ (Step S42). ０ " is "M ｍａｘ " (step S35: Yes), the false authentication rate calculation unit 3122 calculates "P ｓ " is set to "0" (step S43).

[0094] The false authentication rate calculation unit 3122 calculates the number of false acceptances "O Ｓ " to "P Ｓ " are added together (step S44). The false authentication rate calculation unit 3122 determines whether L is less than K (step S34). If L is less than K (step S34: Yes), the process returns to step S31, and the next process using the past authentication result is performed. If there are no unprocessed past authentication results (step S34: No), the false authentication rate calculation unit 3122 calculates the false acceptance rate of falsely accepting a different person at a predetermined threshold value of the matching score (step S45). The false authentication rate calculation unit 3122 calculates the false acceptance rate of falsely accepting a different person at a predetermined threshold value of the matching score (step S46). Ｓ Calculate the false acceptance rate by dividing " by "K".

[0095] That is, the false authentication rate calculation unit 3122 can estimate the false acceptance rate of falsely accepting a person at a predetermined threshold value of the matching score using the false authentication success rate. [3-4: Calculation of false rejection rate and false acceptance rate performed by the information processing device 3]

[0096] Next, a calculation operation when both rejection by the person in question and acceptance by others are taken into consideration during authentication processing will be described.

[0097] As shown in FIG. 10, the false authentication rate calculation unit 3122 calculates the false authentication rate "L" and the number of false rejections "S ｆ " and the number of times of accepting others "O Ｓ " to "0" (step S30). The false authentication rate calculation unit 3122 acquires past authentication results from the authentication result storage unit 322 (step S31). The false authentication rate calculation unit 3122 increments "L" by "1" (step S32).

[0098] The false authentication rate calculation unit 3122 determines whether the final true person matching score included in the past authentication result exceeds a predetermined threshold or whether the final other person matching score exceeds a predetermined threshold (step S46). If the final true person matching score exceeds a predetermined threshold or the final other person matching score exceeds a predetermined threshold (step S46: Yes), the false authentication rate calculation unit 3122 determines whether the final other person matching score exceeds a predetermined threshold (step S40). If the final other person matching score exceeds a predetermined threshold (step S40: Yes), the false authentication rate calculation unit 3122 determines whether the final other person matching score exceeds a predetermined threshold (step S40: Yes). If the final other person matching score exceeds a predetermined threshold, the false authentication rate calculation unit 3122 determines whether the false person matching score exceeds a predetermined threshold (step S40: Yes). Ｓ " is incremented by "1" (step S41).

[0099] If the final identity matching score does not exceed the predetermined threshold and the final other person matching score does not exceed the predetermined threshold (step S46: No), the false authentication rate calculation unit 3122 calculates "M ０ " is "M ｍａｘ " (Step S35). ０ " is "M ｍａｘ If "M" is not the case (step S35: No), the probability of failure in identity verification and the probability of success in identity verification when the authentication process is repeated up to the maximum number of times as described with reference to FIG. 6A are calculated (step S47).０ " is "M ｍａｘ ” (step S35: Yes), the authentication success / failure probability calculation unit 2121 calculates the false authentication rate calculation unit 3122 as “P ｆ " to "1" and "P ｓ " is set to "0" (step S48).

[0100] The false authentication rate calculation unit 3122 calculates the number of false rejections "S ｆ " to "P ｆ Furthermore, the false authentication rate calculation unit 3122 calculates the number of false acceptances "O Ｓ " to "P ｓ " are added together (step S49). The false authentication rate calculation unit 3122 determines whether L is less than K (step S34). If L is less than K (step S34: Yes), the process returns to step S31, and the next processing is performed using the past authentication result. If there are no unprocessed past authentication results (step S34: No), the false authentication rate calculation unit 3122 calculates the false rejection rate and false acceptance rate at a predetermined threshold value of the matching score (step S50). The false authentication rate calculation unit 3122 calculates the false rejection rate and false acceptance rate at a predetermined threshold value of the matching score. ｆ ” by “K” to calculate the false rejection rate, and the number of false acceptances “O Ｓ Calculate the false acceptance rate by dividing " by "K".

[0101] The output control unit 213 may cause the output device 25 to output an estimated value of at least one of the false rejection rate and the false acceptance rate according to a predetermined threshold value of the matching score. [3-5: Technical Effects of the Information Processing Device 3]

[0102] The information processing device 3 in the third embodiment can accurately estimate the false rejection rate corresponding to a threshold value based on the probability of failure in identity matching, and can also accurately estimate the false acceptance rate corresponding to a threshold value based on the probability of success in false identity matching. [4: Fourth Embodiment]

[0103] Next, a fourth embodiment of the information processing device, the information processing method, and the recording medium will be described. Hereinafter, the fourth embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 4 to which the fourth embodiment of the information processing device, the information processing method, and the recording medium is applied. [4-1: Configuration of the information processing device 4]

[0104] The configuration of the information processing device 4 in the fourth embodiment will be described with reference to Fig. 11. Fig. 11 is a block diagram showing the configuration of the information processing device 4 in the fourth embodiment.

[0105] 11, the information processing device 4 in the fourth embodiment differs from the information processing device 2 in the second embodiment and the information processing device 3 in the third embodiment in that the estimation unit 412 includes a table generation unit 4123 and a table storage unit 423 is implemented in the storage device 22. Other features of the information processing device 4 may be the same as other features of at least one of the information processing device 2 and the information processing device 3. Therefore, hereinafter, only the parts that differ from the embodiments already described will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate. [4-2: Operation of Calculating the False Authentication Rate for Each Threshold Value Performed by the Information Processing Device 4]

[0106] The table generating unit 4123 gradually changes the threshold value TH of the matching score, and estimates at least one of the false rejection rate of falsely rejecting the correct person and the false acceptance rate of falsely accepting a different person for each changed threshold value.

[0107] The flow of the operation of calculating the false authentication rate for each threshold value performed by the information processing device 4 will be described with reference to Fig. 12. Fig. 12 is a flowchart showing the flow of the operation of calculating the false authentication rate for each threshold value performed by the information processing device 4.

[0108] 12, the table generating unit 4123 sets the threshold value TH to an initial value (step S51). The table generating unit 4123 may set the minimum value of the matching score threshold value that may be used during operation as the initial value of the threshold value TH.

[0109] The false authentication rate calculation unit 3122 calculates at least one of a false rejection rate and a false acceptance rate for the set threshold value TH (step S52). The false authentication rate calculation unit 3122 may calculate at least one of a false rejection rate and a false acceptance rate by performing at least one of the operations described with reference to FIGS.

[0110] The table generation unit 4123 determines whether to calculate at least one of the false rejection rate and the false acceptance rate for the next threshold TH (step S53). The table generation unit 4123 may determine, for example, whether to calculate at least one of the false rejection rate and the false acceptance rate for the next threshold TH, which is larger than the currently set threshold TH by a predetermined change amount ΔTH. The table generation unit 4123 may determine to calculate at least one of the false rejection rate and the false acceptance rate for the next threshold TH, for example, when the next threshold TH is less than the maximum matching score threshold that may be used during operation. If the table generation unit 4123 determines to calculate at least one of the false rejection rate and the false acceptance rate for the next threshold TH (step S53: Yes), the table generation unit 4123 sets the next threshold TH (step S54) and proceeds to step S51.

[0111] If the calculation of at least one of the false rejection rate and the false acceptance rate for the next threshold value TH is not performed (step S55: No), the table generation unit 4123 generates a table including the threshold value TH and at least one of the false rejection rate and the false acceptance rate corresponding to the threshold value TH (step S51). The table generation unit 4123 stores the generated table in the table storage unit 423. [4-3: Output Operation Performed by Information Processing Device 4]

[0112] The output control unit 413 refers to the table stored in the table storage unit 423, and when one of the false rejection rate and the false acceptance rate is specified, outputs the other of the false rejection rate and the false acceptance rate. As illustrated in Fig. 13, when one of the false rejection rate and the false acceptance rate is specified, the output control unit 413 may display the other of the false rejection rate and the false acceptance rate on a display serving as the output device 25. The output control unit 413 may display a threshold value TH corresponding to the specified one of the false rejection rate and the false acceptance rate.

[0113] For example, when a person who sets the threshold value TH of the matching score specifies at least one of a desired false rejection rate and a false acceptance rate, the output control unit 413 may display combinations of false rejection rates, false acceptance rates, and threshold value TH whose values ​​are close to at least one of the desired false rejection rate and false acceptance rate. The output control unit 413 may highlight combinations of false rejection rates, false acceptance rates, and threshold value TH that are determined to be particularly appropriate, as illustrated in FIG. 13. [4-4: Technical Effects of the Information Processing Device 4]

[0114] The information processing device 4 in the fourth embodiment can accurately estimate the corresponding false rejection rate and false acceptance rate for each threshold value TH of the matching score. The information processing device 4 can assist in setting an appropriate threshold value TH of the matching score according to the desired false rejection rate and false acceptance rate. [5: Fifth Embodiment]

[0115] Next, a fifth embodiment of an information processing device, an information processing method, and a recording medium will be described. Hereinafter, the fifth embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 5 to which the fifth embodiment of the information processing device, the information processing method, and the recording medium is applied. [5-1: Configuration of Information Processing Device 5]

[0116] The configuration of the information processing device 5 in the fifth embodiment will be described with reference to Fig. 14. Fig. 14 is a block diagram showing the configuration of the information processing device 5 in the fifth embodiment.

[0117] As shown in Fig. 14, the information processing device 5 in the fifth embodiment differs from the information processing device 2 in the second embodiment to the information processing device 4 in the fourth embodiment in that the estimation unit 512 includes a function generation unit 5124 and a function storage unit 524 is implemented in the storage device 22. Other features of the information processing device 5 may be the same as at least one other feature of the information processing device 2 to the information processing device 4. Therefore, hereinafter, differences from the already-described embodiments will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate. [5-2: Function Generation Operation Performed by Information Processing Device 5]

[0118] The flow of the function generation operation performed by the information processing device 5 will be described with reference to Fig. 15. Fig. 15 is a flowchart showing the flow of the function generation operation performed by the information processing device 5.

[0119] 15 , the function generation unit 5124 sets the threshold value TH to an initial value (step S51). The false authentication rate calculation unit 3122 calculates at least one of the false rejection rate and the false acceptance rate for the set threshold value TH (step S52). If the function generation unit 5124 determines that the false rejection rate and / or the false acceptance rate for the next threshold value TH is to be calculated (step S53: Yes), the function generation unit 5124 sets the next threshold value TH (step S54) and returns to step S51.

[0120] If the calculation of at least one of the false rejection rate and the false acceptance rate for the next threshold value TH is not performed (step S53: No), the function generating unit 5124 generates a function of at least one of the false rejection rate and the false acceptance rate for the threshold value TH (step S60). The function generating unit 5124 may derive the function from the threshold value TH and discrete values ​​of at least one of the false rejection rate and the false acceptance rate corresponding to the threshold value TH. [5-3: Output operation performed by the information processing device 5]

[0121] The output control unit 513 outputs at least one of the false rejection rate and the false acceptance rate as a function of the threshold value TH. The output control unit 513 may perform control so as to output a graph of the function generated by the function generation unit 5124.

[0122] For example, the output control unit 513 may display a graph on a display as the output device 25, as illustrated in FIG. 16, in which the horizontal axis represents the threshold value TH and the vertical axis represents the false rejection rate FNIR and the false acceptance rate FPIR.

[0123] Furthermore, the output control unit 513 may output at least one of the false rejection rate and the false acceptance rate corresponding to the specified threshold value TH in accordance with the designation of the threshold value TH. The output control unit 513 may provide an interface that allows the user to easily recognize the false rejection rate and the false acceptance rate corresponding to a specific threshold value. For example, as illustrated in FIG. 16, the output control unit 513 may provide an interface that allows the user to designate a desired threshold value by moving the position of the arrow A with a slider S, and displays the numerical values ​​of the false rejection rate and the false acceptance rate corresponding to the specified specific threshold value. FIG. 16 illustrates a case where the false rejection rate corresponding to the specified threshold value is "2.8%" and the false acceptance rate corresponding to the specified threshold value is "0.5%". [5-4: Technical Effects of the Information Processing Device 5]

[0124] The information processing device 5 according to the fifth embodiment can accurately estimate the false rejection rate and false acceptance rate corresponding to the threshold value TH of the matching score, and can present the false rejection rate and false acceptance rate corresponding to the threshold value TH in an easily understandable manner. [6: Sixth Embodiment]

[0125] Next, a sixth embodiment of an information processing device, an information processing method, and a recording medium will be described. Hereinafter, the sixth embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 6 to which the sixth embodiment of the information processing device, the information processing method, and the recording medium is applied. [6-1: Configuration of Information Processing Device 6]

[0126] The configuration of the information processing device 6 in the sixth embodiment will be described with reference to Fig. 17. Fig. 17 is a block diagram showing the configuration of the information processing device 6 in the sixth embodiment.

[0127] 17, the information processing device 6 in the sixth embodiment differs from the information processing device 2 in the second embodiment to the information processing device 5 in the fifth embodiment in that the estimation unit 612 includes a time-slot estimation unit 6125 and a time-slot false authentication rate storage unit 625 is implemented in the storage device 22. Other features of the information processing device 6 may be the same as at least one other feature of the information processing device 2 to the information processing device 5. Therefore, hereinafter, only the parts that differ from the already-described embodiments will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate. [6-2: Operations Performed by Information Processing Device 6]

[0128] The authentication result storage unit 622 in this embodiment may include information indicating the time period in which the authentication process was performed in the authentication result, or may store the authentication result for each time period.

[0129] The time-slot estimation unit 6125 estimates at least one of the false rejection rate and the false acceptance rate for each time slot in which the authentication process is performed. That is, the K authentication results described in the third embodiment may be divided into groups for each time slot, and the false rejection rate and the false acceptance rate may be calculated for each group. The time-slot estimation unit 6125 may gradually change the threshold value TH for each time slot, and estimate at least one of the false rejection rate and the false acceptance rate for each changed threshold value TH.

[0130] That is, the estimation unit 612 may perform, for each time period during which the authentication process is performed, the operation of estimating at least one of the probability of failure in identity matching and the probability of success in false matching, as described with reference to Figures 5 and 6, and the operation of estimating at least one of the false rejection rate and false acceptance rate, as described with reference to Figures 8 to 10, while changing the threshold value as shown in Figures 12 and 15.

[0131] The time zone estimation unit 6125 may use both the actual measured values ​​stored in the authentication result storage unit 622 and estimated values ​​estimated based on the actual measured values ​​to determine the trends in at least one of the false rejection rate and false acceptance rate for each time zone.

[0132] The time zone estimation unit 6125 may gradually change the threshold value TH for each time zone, and store at least one of the false rejection rate and the false acceptance rate estimated for each changed threshold value TH in the time zone false authentication rate storage unit 625.

[0133] The output control unit 613 outputs at least one of the false rejection rate and the false acceptance rate for each time period in which the authentication process was performed in a visually recognizable manner. The output control unit 613 may refer to the time period false authentication rate storage unit 625 and cause the display serving as the output device 25 to display the false rejection rate and the false acceptance rate for each time period.

[0134] The output control unit 613 may graph at least one of the false rejection rate and the false acceptance rate for each time period, and output information that allows the user to easily understand the changes for each time period. For example, as illustrated in FIG. 18( a), the output control unit 613 may display the false rejection rate FNIR corresponding to the time period when the threshold value TH is fixed. The output control unit 613 may display the false acceptance rate FPIR corresponding to the time period when the threshold value TH is fixed. FIG. 18( a) illustrates that the false rejection rate FNIR tends to be higher in the evening hours when the setting sun has an effect. In other words, the output control unit 613 can allow the user to understand that the false rejection rate FNIR tends to be higher in the evening hours when the setting sun has an effect.

[0135] 18(b), the output control unit 613 may display the threshold value TH corresponding to the time period when the false rejection rate FNIR is fixed. The output control unit 613 may also display the threshold value TH corresponding to the time period when the false acceptance rate FPIR is fixed. The output control unit 613 may output information that enables the user to recognize the transition of the threshold value TH for each time period, thereby assisting the user in determining the threshold value for the matching score.

[0136] Although the description has been given of changes according to time zones on a daily basis, the estimation unit 612 may estimate changes according to, for example, weekly, monthly, or yearly units. Also, estimation may be performed by distinguishing time zones in consideration of seasonal changes. For example, the length of daylight differs between summer and winter, so the tendency may change even during the same time zone of the day. Therefore, estimation may be performed by distinguishing time zones by season. [6-3: Technical Effects of the Information Processing Device 6]

[0137] The information processing device 6 in the sixth embodiment can accurately estimate at least one of the false rejection rate and the false acceptance rate for each time period. The information processing device 6 presents at least one of the false rejection rate and the false acceptance rate for each time period, thereby assisting in setting a threshold value for the matching score according to the time period. [7: Seventh Embodiment]

[0138] Next, a seventh embodiment of an information processing device, an information processing method, and a recording medium will be described. Hereinafter, the seventh embodiment of the information processing device, the information processing method, and the recording medium will be described using an information processing device 7 to which the seventh embodiment of the information processing device, the information processing method, and the recording medium is applied. [7-1: Configuration of Information Processing Device 7]

[0139] The configuration of the information processing device 7 in the seventh embodiment will be described with reference to Fig. 19. Fig. 19 is a block diagram showing the configuration of the information processing device 7 in the seventh embodiment.

[0140] As shown in Fig. 19, the information processing device 7 in the seventh embodiment differs from the information processing device 2 in the second embodiment to the information processing device 6 in the sixth embodiment in that an occurrence probability estimation model generation unit 714 is implemented in the calculation device 21. Other features of the information processing device 7 may be the same as at least one other feature of the information processing device 2 to the information processing device 6. Therefore, in the following, differences from the already-described embodiments will be described in detail, and descriptions of other overlapping parts will be omitted as appropriate. [7-2: Occurrence Probability Estimation Model]

[0141] The occurrence probability estimation model is a model that, when a matching score generated in a past authentication process by the authentication unit 211 is input, outputs the occurrence probability of the matching score that would be generated if the authentication process were performed in a corresponding session after the past authentication process.

[0142] The occurrence probability estimation model is a model specific to the authentication mechanism used by the authentication unit 211. The occurrence probability estimation model is generated from the authentication result of biometric authentication using the authentication mechanism and is prepared in advance. [7-3: Generation of occurrence probability estimation model performed by the information processing device 7]

[0143] The occurrence probability estimation model generation unit 714 generates an occurrence probability estimation model. The occurrence probability estimation model generation unit 714 generates the occurrence probability estimation model based on past authentication results by the authentication unit 211. The occurrence probability estimation model generation unit 714 may generate the occurrence probability estimation model based on past authentication results by the authentication unit 211 stored in the authentication result storage unit 722. The past authentication results stored in the authentication result storage unit 722 may include matching scores of all repeated authentication processes including the final one.

[0144] The authentication mechanism may, for example, authenticate an object approaching the authentication mechanism. In this case, since the object gradually approaches the authentication mechanism, it is expected that the accuracy of the acquired biometric information will improve with each repetition of the authentication process. With each repetition of the authentication process, the image will become clearer, and it is expected that an image more suitable for biometric authentication will be acquired. Therefore, it is expected that the matching score will increase with each repetition. For this reason, for example, when authenticating an object approaching the authentication mechanism, the occurrence probability of the next matching score may depend on the magnitude of the previous matching score.

[0145] For example, the authentication results obtained by carrying out multiple authentication processes by changing the distance between the subject and the camera may be stored in the authentication result storage unit 722. This storage may be carried out as an experiment before the operational stage when the authentication mechanism is used for actual biometric authentication. The occurrence probability estimation model generation unit 714 may generate an occurrence probability estimation model based on the authentication results of multiple authentication processes carried out as this experiment. Specifically, the change in score between successive authentications of the same subject is tracked to obtain the distribution of matching scores. The matching score for the current round is calculated as s ｐ The matching score for the next round is s ｎ Then, from the past history, the same s ｐ (s included in the given range ｐ ) for each ｎ The distribution of s is calculated. ｐ A model for each value of .times. ...

[0146] The authentication result storage unit 722 may also store authentication results of authentication processes performed in the operation stage when the authentication mechanism is actually used for biometric authentication. For example, the parameters of the occurrence probability estimation model may be weighted according to the authentication results stored in the operation stage. Since the authentication results stored in the operation stage reflect the situation at the site where the authentication mechanism is used, the occurrence probability estimation model can be adapted to the situation at the site. The situation at the site may include the environment at the site where the authentication is performed, the purpose of the authentication, etc. [7-4: Input to the Occurrence Probability Estimation Model]

[0147] The estimation unit 712 calculates the occurrence probability using an occurrence probability estimation model. The estimation unit 712 inputs a matching score generated in a past authentication process into the occurrence probability estimation model, and outputs the occurrence probability of a matching score that would be generated if an authentication process were performed in a corresponding round after the past authentication process, which is output from the occurrence probability estimation model. For example, the estimation unit 712 inputs an end matching score into the occurrence probability estimation model, and outputs the occurrence probability of a matching score that would be generated if an authentication process were performed the next time after the end round, which is output from the occurrence probability estimation model.

[0148] The occurrence probability estimation model may be input with the matching score of the round before the relevant round after the past authentication process. In other words, when the matching score generated in the current authentication process (the round before the relevant round after the past authentication process) is input, the occurrence probability estimation model may output the occurrence probability of the matching score generated in the next authentication process (the relevant round after the past authentication process). The occurrence probability estimation model may first be input with the ending matching score of the ending round when the iterations of the authentication process have ended.

[0149] Alternatively, a representative score generated based on a plurality of matching scores generated in the authentication processes repeated up to the time before the relevant time after the past authentication process may be input to the occurrence probability estimation model. A representative score (a statistical representative value of the scores) generated based on a plurality of matching scores generated up to the end of the repetition of the authentication process may be input to the occurrence probability estimation model. The representative score may be, for example, an average score obtained by averaging the individual matching scores. Alternatively, the representative score may be the median of the individual matching scores. When the representative score is used as input, it is possible to suppress the influence of a large matching score being generated by chance. [7-5: Occurrence Probability Estimation Model of Identity Verification Score]

[0150] The identity verification score occurrence probability estimation model may output the occurrence probability of an identity verification score to be generated in the next authentication process when the identity verification score generated in the current authentication process is input.

[0151] Alternatively, the model for estimating the probability of occurrence of an identity matching score may output the probability of occurrence of an identity matching score generated in the next authentication process when the representative score described above is input. For example, when photographs are taken multiple times in a stationary state and authentication is performed using the images obtained by the photographs, there is no reason for the matching score to gradually increase, so it does not need to depend on the previous identity matching score. Rather, if the probability of occurrence of the next matching score is calculated from the representative score of the matching scores generated by repeated authentication processes, the accuracy of estimating the probability of occurrence of the matching score may be improved. [7-6: Model for estimating the probability of occurrence of another person's matching score]

[0152] In another embodiment, the case has been described in which it is assumed that the occurrence probability of the other person matching score does not depend on the previous other person matching score, but the occurrence probability of the other person matching score may also depend on the matching score generated in the past authentication process, similar to the occurrence probability of the true person matching score. [7-8: Technical Effects of the Information Processing Device 7]

[0153] The information processing device 7 in the seventh embodiment generates an occurrence probability estimation model and estimates estimated authentication information using the occurrence probability estimation model, so that it can estimate estimated authentication information with high accuracy and can assist in setting an appropriate threshold value for the matching score. The information processing device 7 can generate an occurrence probability estimation model that can output an occurrence distribution of the matching score with high accuracy when it depends on the matching score from the previous time. The information processing device 7 can generate an occurrence probability estimation model that can output an occurrence distribution of the matching score with high accuracy even when it does not depend on the matching score from the previous time. [8: Supplementary Note]

[0154] The following supplementary notes are further disclosed with respect to the above-described embodiments. [Supplementary Note 1] An information processing device comprising: authentication means for performing an authentication process to determine whether authentication of the target will be successful or not based on a matching score generated using biometric information of the target and predetermined conditions; estimation means for estimating estimated authentication information regarding whether authentication of the target will be successful or not when the authentication means repeats the authentication process a predetermined number of times based on past authentication results by the authentication means; and output means for outputting the estimated authentication information. [Supplementary Note 2] The information processing device described in Supplementary Note 1, wherein the predetermined conditions include a condition under which the authentication means determines that authentication of the target has been successful and terminates the repetition of the authentication process when the matching score exceeds a threshold value of the matching score, and a condition under which the authentication means terminates the repetition when it has repeated the authentication process a maximum number of times. [Supplementary Note 3] The information processing device described in Supplementary Note 1 or 2, wherein the past authentication results include a final matching score for a final iteration of the authentication process when the repetition of the authentication process has ended, and the estimation means calculates the occurrence probability of a matching score that would be generated if the authentication process were performed after the final iteration, based on the final matching score. [Supplementary Note 4] The information processing device according to Supplementary Note 3, wherein the estimation means uses the occurrence probability to estimate at least one of an identity verification failure probability indicating the probability of failing to verify identity verification with the person in question when the authentication means repeats the authentication process a predetermined number of times at a predetermined threshold value of the matching score, and a false identity verification success probability indicating the probability of succeeding in verifying identity verification with a person other than the person in question. [Supplementary Note 5] The information processing device according to Supplementary Note 4, wherein the estimation means performs at least one of an operation of estimating a false rejection rate of falsely rejecting the person in question using the identity verification failure probability and an operation of estimating a false acceptance rate of falsely accepting a person in question using the false identity verification success probability at a predetermined threshold value of the matching score. [Supplementary Note 6] The information processing device according to Supplementary Note 5, wherein the estimation means changes the threshold value of the matching score in stages and estimates at least one of the false rejection rate and the false acceptance rate for each changed threshold value.[Supplementary Note 7] The information processing device according to Supplementary Note 5, wherein the output means outputs at least one of the false rejection rate and the false acceptance rate as a function of a threshold value of the matching score, and outputs at least one of the false rejection rate and the false acceptance rate corresponding to the specified threshold value in response to the specified threshold value. [Supplementary Note 8] The information processing device according to Supplementary Note 5, wherein the output means, when one of the false rejection rate and the false acceptance rate is specified, outputs the other of the false rejection rate and the false acceptance rate. [Supplementary Note 9] The information processing device according to Supplementary Note 5, wherein the estimating means estimates at least one of the false rejection rate and the false acceptance rate for each time period in which the authentication process is performed. [Supplementary Note 10] The information processing device according to Supplementary Note 9, wherein the output means outputs at least one of the false rejection rate and the false acceptance rate for each time period in which the authentication process is performed. [Supplementary Note 11] The information processing device according to Supplementary Note 1 or 2, further comprising: a generation means for generating, when inputted with a matching score generated in a past authentication process by the authentication means, an occurrence probability estimation model that outputs the occurrence probability of a matching score generated when the authentication process is performed a corresponding time after the past authentication process, wherein the generation means generates the occurrence probability estimation model based on past authentication results by the authentication means, and the estimation means estimates the occurrence probability using the occurrence probability estimation model. [Supplementary Note 12] The information processing device according to Supplementary Note 11, wherein the matching score generated in the past authentication process by the authentication means is the matching score of a time previous to the corresponding time. [Supplementary Note 13] The information processing device according to Supplementary Note 11, wherein the matching score generated in the past authentication process is a representative score generated based on a plurality of matching scores generated in authentication processes repeated up to the time previous to the corresponding time. [Supplementary Note 14] An information processing method, in which an authentication means performs an authentication process to determine whether authentication of the target will be successful or not based on a matching score generated using the target's biometric information and predetermined conditions, estimates estimated authentication information regarding whether authentication of the target will be successful or not when the authentication means repeats the authentication process a predetermined number of times based on past authentication results by the authentication means, and outputs the estimated authentication information.[Supplementary Note 15] A recording medium having recorded thereon a computer program for causing a computer to execute an information processing method, the method comprising: performing an authentication process in which an authentication means determines whether authentication of the target will be successful or not based on a matching score generated using biometric information of the target and predetermined conditions; estimating estimated authentication information regarding whether authentication of the target will be successful or not when the authentication means repeats the authentication process up to a predetermined number of times based on past authentication results by the authentication means; and outputting the estimated authentication information.

[0155] This disclosure may be modified as appropriate within the scope of the claims and the technical idea that can be read from the entire specification. Information processing devices, information processing methods, and recording media that involve such modifications are also included in the technical idea of ​​this disclosure.

[0156] 1, 2, 3, 4, 5, 6, 7 Information processing device 11, 211 Authentication unit 12, 212, 312, 412, 512, 612, 712 Estimation unit 13 Output unit 213, 413, 513, 613 Output control unit 2111 Acquisition unit 2112 Matching score generation unit 2113 Success / failure determination unit 2121 Authentication success / failure probability calculation unit 221 Biometric information registration unit 222, 322, 622, 722 Authentication result storage unit 3122 False authentication rate calculation unit 4123 Table generation unit 423 Table storage unit 5124 Function generation unit 524 Function storage unit 6125 Time zone estimation unit 714 Occurrence probability estimation model generation unit

Claims

1. an authentication means for determining whether authentication of the target is successful or not based on a result of a matching process that determines whether matching of the target is successful or not based on a matching score generated using the target's biometric information and on predetermined conditions; an estimation means for estimating estimated authentication information relating to success or failure of authentication of the target when the authentication means repeats the matching process a predetermined number of times based on past authentication results by the authentication means; an output means for outputting the estimated authentication information; An information processing device comprising:

2. The predetermined conditions include a condition in which the authentication means determines that the authentication of the target is successful and terminates the repetition of the matching process when the matching score exceeds the matching score threshold, and a condition in which the authentication means terminates the repetition when the matching process is repeated an upper limit number of times. The information processing device according to claim 1 .

3. The past authentication result includes a final matching score at the end of the repetition of the matching process, The estimation means calculates the probability of occurrence of a matching score generated when the matching process is performed after the final time, according to the final matching score.

3. The information processing device according to claim 1.

4. The estimation means uses the occurrence probability to estimate at least one of a false matching failure probability indicating a probability of failing to match the person in question when the authentication means repeats the matching process a predetermined number of times at a predetermined threshold value of the matching score, and a false matching success probability indicating a probability of succeeding in matching with a person other than the person in question. The information processing device according to claim 3 .

5. The estimation means performs at least one of an operation of estimating a false rejection rate of falsely rejecting the person using the false matching failure probability at a predetermined threshold value of the matching score, and an operation of estimating a false acceptance rate of falsely accepting a person using the false matching success probability. The information processing device according to claim 4 .

6. The estimation means changes a threshold value of the matching score in stages and estimates at least one of the false rejection rate and the false acceptance rate for each changed threshold value. The information processing device according to claim 5 .

7. The output means outputting at least one of the false rejection rate and the false acceptance rate as a function of the match score threshold; In response to the designation of the threshold, at least one of the false rejection rate and the false acceptance rate corresponding to the designated threshold is output. The information processing device according to claim 5 .

8. a generation means for generating an occurrence probability estimation model that, when a matching score generated in a past matching process by the authentication means is input, outputs the occurrence probability of a matching score that would be generated if the matching process were performed a corresponding time after the past matching process, the generation means generates the occurrence probability estimation model based on past authentication results by the authentication means; The estimation means estimates the occurrence probability using the occurrence probability estimation model.

3. The information processing device according to claim 1.

9. Determining whether authentication of the target has been successful based on a result of a matching process that determines whether matching of the target has been successful based on a matching score generated using the target's biometric information and on predetermined conditions; based on past authentication results by the authentication means, estimating estimated authentication information regarding success or failure of authentication of the target when the authentication means repeats the matching process a predetermined number of times; Outputting the estimated authentication information A computer-implemented information processing method.

10. On the computer, determining whether authentication of the target has been successful based on a result of a matching process that determines whether matching of the target has been successful based on a matching score generated using the target's biometric information and on predetermined conditions; based on past authentication results by the authentication means, estimating estimated authentication information regarding success or failure of authentication of the target when the authentication means repeats the matching process a predetermined number of times; Outputting the estimated authentication information A computer program for executing an information processing method.