Information processor, information processing method, and program

Abstract

To provide an information processor, an information processing method, and a program enabling robust authentication using voice of a target.SOLUTION: An information processor includes an output part, a voice acquisition part, and an authentication-related processing part. The output part outputs output information containing randomly generated random information. The voice acquisition part acquires response voice for the output information. The authentication-related processing part performs processing related to authentication of a target using at least one of voiceprint information and word information contained in the response voice.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] For example, the authentication device described in Patent Document 1 includes a biometric authentication unit, a question authentication unit, a score calculation unit, a determination unit, a biometric authentication information storage unit, and a question authentication information storage unit.

[0003] The biometric authentication unit inputs biometric authentication information of the person to be authenticated, compares it with the biometric authentication information stored in the biometric authentication information storage unit, quantitatively determines the identity of the person, and assigns a biometric authentication score.

[0004] The question authentication unit presents a question to the person to be authenticated and allows them to input an answer, determines whether the input answer matches a pre-registered answer for the question, and assigns a question authentication score depending on whether the answer is correct or incorrect. The question authentication unit associates scores assigned between 0 and 50 depending on the difficulty of the question with the question and the answer and stores them in the question authentication information storage unit, and if the input answer matches the registered answer, assigns the score associated with the question as the question authentication score. If they do not match, that is, if the answer is incorrect, the question authentication unit sets the question authentication score to 0. The question authentication unit assigns low scores to questions with low difficulty and high scores to questions with high difficulty.

[0005] The score calculation unit receives the question authentication score from the question authentication unit and calculates a total score by adding up the biometric authentication score and the question authentication score. The determination unit determines whether the total score is equal to or greater than a predetermined set value, and determines that authentication is successful if the total score is equal to or greater than the set value, or determines that authentication is unsuccessful if the total score is lower than the set value and the number of questions asked by the question authentication unit has reached a predetermined number.

[0006] The biometric authentication information storage unit stores the biometric authentication information of each person to be authenticated in association with a user ID that identifies the person to be authenticated. The question authentication information storage unit stores multiple pairs of questions, answers, and scores for each person to be authenticated. The person to be authenticated registers answers to questions prepared in advance in the question authentication information storage unit. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-85913 Summary of the Invention [Problem to be solved by the invention]

[0008] The present disclosure aims to improve upon the techniques described in the prior art documents mentioned above. [Means for solving the problem]

[0009] The information processing device according to the present disclosure includes: an output means for outputting output information including randomly generated random information; a voice acquisition means for acquiring a response voice in response to the output information; and an authentication-related processing means for performing processing related to the authentication of a target using at least one of voiceprint information and word information included in the response voice.

[0010] The information processing method in the present disclosure includes: One or more computers Outputting output information including randomly generated random information; acquiring a response voice in response to the output information; At least one of voiceprint information and word information contained in the response voice is used to perform processing related to authentication of the target.

[0011] The program in this disclosure is On one or more computers, Outputting output information including randomly generated random information; acquiring a response voice in response to the output information; The program is for executing a process related to authentication of a target using at least one of voiceprint information and word information contained in the response voice. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a block diagram showing a configuration example of a first information processing device according to the present disclosure. [Figure 2] 1 is a diagram illustrating a configuration example of a first information processing system according to the present disclosure. [Figure 3] 10 is a flowchart illustrating an example of a first information process according to the present disclosure. [Figure 4] FIG. 2 is a diagram illustrating a configuration example of a first terminal according to the present disclosure. [Figure 5] FIG. 2 is a diagram illustrating a detailed configuration example of a first information processing device according to the present disclosure. [Figure 6] 10 is a flowchart showing a detailed example of first information processing according to the present disclosure. [Figure 7] 1 is a block diagram illustrating an example of the physical configuration of a first information processing device according to the present disclosure. [Figure 8] FIG. 2 is a diagram illustrating a configuration example of a second information processing device according to the present disclosure. [Figure 9] FIG. 10 is a diagram illustrating a configuration example of a second registration unit according to the present disclosure. [Figure 10] FIG. 10 is a diagram illustrating a configuration example of a second authentication unit according to the present disclosure. [Figure 11] 10 is a flowchart illustrating an example of a first registration process according to the present disclosure. [Figure 12] 10 is a flowchart illustrating an example of a first authentication process according to the present disclosure. [Figure 13] FIG. 10 is a diagram illustrating a configuration example of a third information processing device according to the present disclosure. [Figure 14] FIG. 10 is a diagram illustrating a configuration example of a third registration unit according to the present disclosure. [Figure 15]FIG. 10 is a diagram illustrating a configuration example of a third authentication unit according to the present disclosure. [Figure 16] 10 is a flowchart illustrating an example of a second registration process according to the present disclosure. [Figure 17] 10 is a flowchart illustrating an example of a second authentication process according to the present disclosure. [Figure 18] FIG. 10 is a diagram illustrating a configuration example of a fourth information processing device according to the present disclosure. [Figure 19] FIG. 10 is a diagram illustrating a configuration example of a fourth registration unit according to the present disclosure. [Figure 20] FIG. 10 is a diagram illustrating a configuration example of a fourth authentication unit according to the present disclosure. [Figure 21] 10 is a flowchart illustrating an example of a third registration process according to the present disclosure. [Figure 22] 10 is a flowchart illustrating an example of a third authentication process according to the present disclosure. [Figure 23] FIG. 10 is a diagram illustrating a configuration example of a fifth information processing device according to the present disclosure. [Figure 24] FIG. 10 is a diagram illustrating a configuration example of a fifth registration unit according to the present disclosure. [Figure 25] FIG. 10 is a diagram illustrating a configuration example of a fifth authentication unit according to the present disclosure. [Figure 26] 10 is a flowchart showing an example of a fourth registration process according to the present disclosure. [Figure 27] 10 is a flowchart showing an example of a fourth authentication process according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In all drawings, similar components are denoted by similar reference numerals, and descriptions thereof will be omitted as appropriate. In the present disclosure, the drawings relate to one or more embodiments.

[0014] [Embodiment 1] (overview) In the technology described in Patent Document 1, in order to achieve high-precision authentication, authentication is performed using biometric authentication information and answers to questions prepared in advance. However, the answers are to the questions prepared in advance. Therefore, there is a risk that a third party may prepare answers to the questions prepared in advance by, for example, investigating the person to be authenticated, and impersonate the person to be authenticated.

[0015] One of the objectives of this disclosure is to enable strong authentication using a subject's voice.

[0016] The information processing device 100 includes an output unit 130, a voice acquisition unit 150, and an authentication-related processing unit 170, as shown in FIG.

[0017] The output unit 130 outputs output information including randomly generated random information.

[0018] The voice acquisition unit 150 acquires a response voice in response to the output information.

[0019] The authentication-related processing unit 170 performs processing related to the authentication of the target using at least one of the voiceprint information and word information contained in the response voice.

[0020] According to this information processing device 100, at least one of voiceprint information and word information contained in a response voice to randomly generated random information can be used to authenticate a target.

[0021] For example, if the random information is information that requires unfamiliar random utterance, unlike familiar utterances such as common phrases, the response voice is likely to be halting and unnatural. Therefore, the difference in voice characteristics between the person in question and a third party impersonating the person in question is likely to be large, making it difficult for a third party to impersonate the person in question.

[0022] For example, if the random information requires a random response, it is more difficult for a third party attempting to impersonate the person to prepare a response to the random information than to prepare a response to fixed information. This makes it more difficult for a third party to impersonate the person.

[0023] In this way, the use of random information makes it difficult for a third party to impersonate someone, thereby enabling strong authentication using the target's voice.

[0024] 2, the information processing system SYS includes an information processing device 100 that performs processing related to authentication of a target, and a terminal 50 that the target uses for a call. The terminal 50 includes a call unit 51 that transmits a response voice. The output unit 130 outputs output information to the terminal 50, for example. The voice acquisition unit 150 acquires the response voice from the terminal 50, for example.

[0025] As described above, the information processing system SYS uses random information to make it difficult for a third party to pose as a person, thereby enabling robust authentication using the target's voice.

[0026] The information processing device 100 executes information processing such as that shown in FIG.

[0027] The output unit 130 outputs output information including the randomly generated random information (step S130).

[0028] The voice acquisition unit 150 acquires a response voice in response to the output information (step S150).

[0029] The authentication-related processing unit 170 performs processing related to the authentication of the target (authentication-related processing) using at least one of the voiceprint information and the word information included in the response voice (step S170).

[0030] According to this information processing, as described above, the use of random information makes it difficult for a third party to impersonate a person, thereby enabling strong authentication using the target's voice.

[0031] (detail) A detailed example of the information processing system SYS will be described below.

[0032] The information processing system SYS is a system that performs processing related to authentication of an object using, for example, voice. The information processing system SYS includes, for example, the terminal 50 and the information processing device 100, as described above.

[0033] The terminal 50 and the information processing device 100 are connected to each other via a communication network NT that may be wired, wireless, or a combination of these, and transmit and receive information to and from each other via the communication network NT.

[0034] The terminal 50 and the information processing device 100 may be connected to each other via a telephone line in addition to or instead of the communication network NT, and may transmit and receive audio via the telephone line.

[0035] (Example of the configuration of the terminal 50) The terminal 50 is a device used by the subject for communication. The terminal 50 may further include a display function for displaying information to the subject, an input function for the subject to input information, and the like.

[0036] As shown in detail in FIG. 4, the terminal 50 includes a call unit 51, a display control unit 53, a display unit 55, an input unit 57, and a terminal communication unit 59.

[0037] The call unit 51 provides, for example, a function for a target to use voice to make a call with the information processing device 100. In detail, for example, the call unit 51 transmits the voice of the target to the information processing device 100, receives voice generated from the information processing device 100, and emits the voice to the target.

[0038] The display control unit 53 displays various types of information on the display unit 55. Under the control of the display control unit 53, the display unit 55 displays various types of information.

[0039] The input unit 57 accepts input of objects including letters, numbers, symbols, and the like.

[0040] The device communication unit 59 communicates information with the information processing device 100. For example, the device communication unit 59 receives display information to be displayed on the display unit 55 from the information processing device 100. For example, the device communication unit 59 transmits information received by the input unit 57 from the user to the information processing device 100 based on an instruction from the user. The information received from the user includes, for example, at least one of letters, numbers, symbols, etc.

[0041] The terminal 50 is, for example, a smartphone or a tablet terminal, but is not limited to these. The terminal 50 may also be, for example, a communication device such as a landline telephone. If the communication device does not have a display function or an input function, the terminal 50 may be used together with the communication device for display or input.

[0042] (Configuration example of information processing device 100) The information processing device 100 is a device that performs processing related to authentication of an object using voice. Processing related to authentication of an object will also be referred to as "authentication-related processing" hereinafter.

[0043] The authentication-related process includes, for example, at least one of a registration process for registering the target and an authentication process for authenticating the target. In the registration process, for example, registration information of the target is initially registered. In the authentication process, for example, the registered target is authenticated using the registration information. The authentication process may include a process for adding registration information of the target, as described below. This added registration information may be used, for example, to authenticate the target in the next authentication process.

[0044] The registration information may be, for example, information that associates personal information, random information (described in detail below), etc. with each target. The personal information is, for example, information that can be used to identify the target.

[0045] The personal information may include, for example, voiceprint information, or may include target information other than voiceprint information for identifying the target.

[0046] The voiceprint information is, for example, information including a feature quantity specific to the target utterance, and includes, for example, one or more values ​​(vectors, etc.).

[0047] The target information may include, for example, one or more of a target ID (Identification), a password, a name, an address, a date of birth, an email address, a telephone number, and biometric information other than fingerprint information. The biometric information other than fingerprint information may be, for example, information about a face, a fingerprint, a palm print, an iris, a vein, etc. This biometric information may include, for example, an image, and may include feature amounts (e.g., a feature vector) extracted from the image.

[0048] The information processing device 100 includes a random information acquisition unit 110, the above-mentioned output unit 130, a voice acquisition unit 150, and an authentication-related processing unit 170, as shown in FIG.

[0049] The random information acquisition unit 110 acquires random information.

[0050] The information processing device 100 executes information processing such as that shown in FIG.

[0051] The random information acquisition unit 110 acquires random information (step S110).

[0052] The above-described steps S130, S150 and S170 are performed.

[0053] (Regarding the random information acquisition unit 110) The random information acquisition unit 110 acquires random information. The random information is randomly generated information. For example, the random information includes at least one of a second question that is randomly generated in association with a first question and a randomly generated character string.

[0054] (Random Information Example 1: Example including questions) When the random information includes a question (second question), the random information acquisition unit 110 may include a generation AI (Artificial Intelligence). For example, when a prompt (command statement) is input, the generation AI outputs a sentence corresponding to the prompt. For example, when a prompt to create a question related to the first question is input, the generation AI outputs a second question. This prompt may include, for example, at least one of the keywords included in the first question and the answer (speech response) to the first question. This generates, for example, a second question related to the keywords included in the first question or the answer (speech response) to the first question.

[0055] A question (second question) related to the first question may be a question that delves deeper into a keyword contained in the first question or the answer (speech response) to the first question. Drilling deeper means seeking more detailed information about the keyword. In this case, the generation AI may be input with, for example, at least one of the first question and a keyword contained in the answer to the first question, as well as a prompt to create a question that seeks more detailed information about the keyword.

[0056] The first question may be, for example, a question prepared in advance, a random question generated using a generation AI, or a question selected from a plurality of candidate questions prepared in advance.

[0057] If the first question is a random question, the first question may be a question output from the generation AI after inputting a prompt to the generation AI to generate a question related to one or more keywords. The one or more keywords may be prepared in advance, selected from multiple keyword candidates, or generated from some or all of the target's personal information (e.g., address).

[0058] When selecting the first question or keyword, this selection may be made by a person such as an operator or a subject, or may be made automatically. When selecting automatically, the first question or keyword may be selected randomly from a plurality of question candidates or a plurality of keyword candidates. Any appropriate method may be used to randomly select the first question or keyword, and for example, a random number N may be used.

[0059] The random number N may be a pseudo-random number or a true random number. The random number N may be generated by, for example, a random number generator. The random number generator may be included in, for example, the random information acquisition unit 110. The random number generator may be provided in, for example, an external device (not shown), and the random information acquisition unit 110 may acquire the random number N from the external device.

[0060] For example, when there are M question candidates (M is an integer equal to or greater than 2), the random information acquisition unit 110 may calculate the remainder Re by dividing the random number N by the number M of question candidates, and set the Rth question candidate output as the first question. Instead of the question candidate and the first question, respectively, a keyword candidate and a keyword may be used.

[0061] The generation AI may output multiple second question candidates for the first question. In this case, the random information acquisition unit 110 may select, for example, a question randomly from the multiple second question candidates as the second question.

[0062] Any appropriate method may be used for the random selection here, and for example, the random number N may be used as described above. For example, when there are M second question candidates, the random information acquisition unit 110 may calculate the remainder Re when dividing the random number N by the number M of second question candidates, and may select the Rth output second question candidate as the second question.

[0063] The random information acquisition unit 110 may transmit the first question to an external device (not shown) equipped with a generation AI, and acquire the second question generated in response thereto from the external device. The method for generating or acquiring a random question is not limited to that described here, and any general method may be used. The second question is not limited to a question related to the first question, and may be a random question different from the first question. A random question different from the first question may be created, for example, by inputting a prompt to the generation AI to create a question different from the first question.

[0064] (Random information example 2: Example including a string) When the random information includes a randomly generated string of characters, the string of characters may include one or more of alphabets, hiragana, katakana, numbers, symbols, etc. The length of the string of characters may be predetermined or may be determined randomly. When a random string of characters is used, the length of the string of characters may be determined randomly between a predetermined minimum length and a predetermined maximum length.

[0065] To generate a random character string, the random information acquisition unit 110 may use, for example, a random number R generated by a random number generator. For example, the random information acquisition unit 110 may obtain a random integer from the random number R for each character constituting the character string and convert this integer into the character.

[0066] The random number R may be a pseudo-random number or a true random number. The random number generator may be included in, for example, the random information acquisition unit 110. The random number generator may be provided in, for example, an external device (not shown), and the random information acquisition unit 110 may acquire the random number R from the external device.

[0067] The method for generating or obtaining a random character string is not limited to the method described here, and any general method may be used.

[0068] (Regarding the output unit 130) The output unit 130 outputs the output information. The output information may include the random information acquired by the random information acquisition unit 110, for example.

[0069] The output information may include, for example, at least one of registration output information and authentication output information. The registration output information is output information that is output for registering an object, and is, for example, output information that is output by the output unit 130 in the registration process. The authentication output information is output information that is output for authenticating an object, and is, for example, output information that is output by the output unit 130 in the authentication process.

[0070] For example, when the output unit 130 outputs either the registration output information or the authentication output information, the one of the output information may include random information. For example, when the output unit 130 outputs the registration output information and the authentication output information, at least one of the registration output information and the authentication output information may include random information.

[0071] The output information may further include a request for predetermined information, which may be, for example, the target information described above.

[0072] Furthermore, the output unit 130 may output the output information to, for example, the terminal 50. The output information may include, for example, audio information for causing the call unit 51 to issue a question or the like, or may include text information for causing the display unit 55 to display a question, a character string, or the like.

[0073] (Regarding the voice acquisition unit 150) The voice acquiring unit 150 acquires, for example, a response voice in response to the output information output by the output unit 130. The voice acquiring unit 150 may acquire the response voice from the call unit 51, for example.

[0074] The response voice may include, for example, at least one of a registration response voice and an authentication response voice. The registration response voice is a response voice to the registration output information, and is, for example, a response voice acquired by the voice acquisition unit 150 in the registration process. The authentication response voice is a response voice to the authentication output information, and is, for example, a response voice acquired by the voice acquisition unit 150 in the authentication process.

[0075] (Regarding the authentication-related processing unit 170) The authentication-related processing unit 170 performs authentication-related processing using at least one of voiceprint information and word information contained in the response voice.

[0076] The word information is, for example, information about phrases included in the response voice, and may include phrases, keywords, character strings, etc. included in the response voice. In detail, for example, the word information may include at least one of a keyword included in the response voice to a question and a randomly generated character string. The question may include a first question and a second question. The keyword is one or more phrases selected from phrases included in the response voice, and may be, for example, a noun, proper noun, etc. included in the response voice. This response voice may include, for example, an answer to the question. The keyword may be, for example, a noun, proper noun, etc. included in the answer. Note that the keyword is not limited to a noun or proper noun.

[0077] The authentication-related processing unit 170 may perform authentication-related processing using, for example, voiceprint information and word information included in the response voice. This allows for two-factor authentication using the response voice and the voiceprint and words, thereby enabling stronger authentication using the target voice.

[0078] As described above, the authentication-related process includes at least one of a registration process for registering a target and an authentication process for authenticating the target. When two-factor authentication is performed, the registration process may include, for example, registering voiceprint information and word information included in the registration response voice as registration information of the target. For example, the authentication process may include authenticating the target using the voiceprint information and word information included in the authentication response voice and the registration information.

[0079] Furthermore, as described above, when the information processing device 100 acquires target information by a method such as including a request for target information in output information, the information processing device 100 may perform authentication using the target information and authentication using voiceprint information and word information. In this case, the information processing device 100 may perform authentication using the target information (first-stage authentication) and authentication using voiceprint information and word information (second-stage authentication). This allows for multi-stage authentication. Therefore, stronger authentication using the target voice is possible.

[0080] Furthermore, when acquiring target information using a response voice, the response voice of the target information can be acquired in addition to the response voice to random information. Therefore, voiceprint information can be obtained from a longer response voice than when only the response voice to random information is acquired, and voiceprint authentication can be performed. Generally, in voiceprint authentication, authentication accuracy is often improved when a relatively long voice is used compared to when a short voice is used. Therefore, more robust authentication using the target voice is possible.

[0081] Here, a general technique may be used to obtain word information or voiceprint information from the response voice.

[0082] For example, authentication-related processing unit 170 may acquire word information from the response voice using a word estimation model, which is a machine learning model for acquiring word information from voice. In this case, for example, authentication-related processing unit 170 may include a word estimation model and acquire word information from the response voice by inputting the response voice into the word estimation model. Also, for example, authentication-related processing unit 170 may transmit the response voice to an external device (not shown) including the word estimation model, and acquire word information acquired in the external device using the word estimation model from the external device. The word estimation model may be, for example, a machine learning model trained using training data including one or more voices and ground truth data indicating word information contained in the voices.

[0083] For example, authentication-related processing unit 170 may acquire voiceprint information from the response voice using a voiceprint estimation model, which is a machine learning model for acquiring voiceprint information from voice. In this case, for example, authentication-related processing unit 170 may include a voiceprint estimation model and acquire voiceprint information from the response voice by inputting the response voice into the voiceprint estimation model. Also, for example, authentication-related processing unit 170 may transmit the response voice to an external device (not shown) including the voiceprint estimation model, and acquire voiceprint information acquired in the external device using the voiceprint estimation model from the external device. The voiceprint estimation model may be, for example, a machine learning model trained using training data including the voices of one or more speakers and correct labels indicating the speakers.

[0084] The machine learning model may be configured using, for example, a neural network, but is not limited to this.

[0085] (Example of physical configuration of information processing device 100) The information processing device 100 physically includes a bus 1010, a processor 1020, a memory 1030, a storage device 1040, a network interface 1050, an input interface 1060, and an output interface 1070, as shown in FIG.

[0086] The bus 1010 is a data transmission path for transmitting and receiving data among the processor 1020, memory 1030, storage device 1040, network interface 1050, input interface 1060, and output interface 1070. However, the method of connecting the processor 1020 and the like to each other is not limited to bus connection.

[0087] The processor 1020 is implemented by a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or the like.

[0088] The memory 1030 is a main storage device realized by a RAM (Random Access Memory) or the like.

[0089] The storage device 1040 is an auxiliary storage device realized by a hard disk drive (HDD), a solid state drive (SSD), a memory card, a read only memory (ROM), etc. The storage device 1040 stores program modules for realizing the functions of the information processing device 100. The processor 1020 reads each of these program modules into the memory 1030 and executes them, thereby realizing the function corresponding to the program module.

[0090] The network interface 1050 is an interface for connecting the information processing device 100 to the communication network NT.

[0091] The input interface 1060 is an interface for the user to input information, and may be configured with, for example, a touch panel, a keyboard, a mouse, a microphone, etc.

[0092] The output interface 1070 is an interface for presenting information to the user, and is configured, for example, by a liquid crystal panel, an organic EL (Electro-Luminescence) panel, etc. The output interface 1070 may also include a speaker.

[0093] Although the information processing device 100 has been described as being physically composed of one device (e.g., a computer), the information processing device 100 may be composed of multiple devices (e.g., computers) that transmit and receive information to and from each other via a communication network NT. In this case, the multiple devices may cooperate to execute information processing.

[0094] (Example of physical configuration of terminal 50) The terminal 50 may be physically configured in the same way as the information processing device 100, or may include an interface for connecting to a telephone line in addition to the configuration of the information processing device 100. Also, the target may use the information processing device 100. In this case, for example, instead of the information processing system SYS including the terminal 50, the information processing device 100 may include the functions of the terminal 50.

[0095] (Actions and Effects) As described above, according to this embodiment, the information processing device 100 includes the output unit 130, the voice acquisition unit 150, and the authentication-related processing unit 170. The output unit 130 outputs output information including randomly generated random information. The voice acquisition unit 150 acquires a response voice in response to the output information. The authentication-related processing unit 170 performs processing related to the authentication of a target using at least one of voiceprint information and word information included in the response voice.

[0096] As a result, as described above, the use of random information makes it difficult for a third party to impersonate a person, thereby enabling robust authentication using the target's voice.

[0097] According to this embodiment, the information processing device 100 further includes a random information acquisition unit 110 that acquires random information. The random information includes at least one of a second question randomly generated in relation to the first question and a randomly generated character string. The word information includes at least one of a keyword included in a response voice to the second question and a randomly generated character string.

[0098] This makes it difficult for a third party to impersonate someone else by using random information, thereby enabling robust authentication using the target's voice.

[0099] According to this embodiment, it further includes a request for predetermined information.

[0100] This allows for the acquisition of a response voice in response to both random information and a request for predetermined information, and therefore allows for the acquisition of a response voice that is longer than the response voice in response to either the random information or a request for predetermined information. Generally, in voice authentication, using a voice that is somewhat longer can improve the accuracy of authentication. Therefore, robust authentication using the target voice is possible.

[0101] According to this embodiment, the authentication-related processing unit 170 performs processing related to the authentication of the target using the voiceprint information and word information contained in the response voice.

[0102] This allows for two-factor authentication using voiceprints and phrases, enabling stronger authentication using the target's voice.

[0103] According to this embodiment, the output information includes registration output information that is output to register the target. The response voice includes a registration response voice that is a response voice to the registration output information. The processing related to authenticating the target includes registering voiceprint information and word information included in the registration response voice as registration information of the target.

[0104] This allows for easy generation and registration of registration information including voiceprint information and word information. Two-factor authentication using voiceprints and words can then be performed. This enables stronger authentication using the target's voice.

[0105] [Embodiment 2] In the second embodiment, an example will be described in which the random information includes a randomly generated character string. In the present embodiment, an example will be described in which the word information includes a character string included in a response voice in response to the random information, i.e., a randomly generated character string. In the present embodiment, for the sake of brevity, descriptions that overlap with other embodiments will be omitted as appropriate.

[0106] (Configuration example of information processing device 200) The information processing system SYS may include an information processing device 200 instead of the information processing device 100. The information processing device 200 includes, for example, a target information acquisition unit 290, a registration information storage unit 295, a random information acquisition unit 210, an output unit 230, a voice acquisition unit 250, and an authentication-related processing unit 270, as shown in FIG.

[0107] The target information acquisition unit 290 acquires target information for identifying a target.

[0108] The registration information storage unit 295 is a storage unit for storing registration information of a target. The registration information may be information that associates the target's voiceprint information with a randomly generated character string. The registration information may also be information that further associates the target's target information.

[0109] The random information acquisition unit 210 acquires random information that includes a randomly generated character string.

[0110] The output unit 230 outputs output information. The output information includes, for example, output information for registration and output information for authentication. For example, the output information for registration includes, as random information, a randomly generated character string. For example, the output information for authentication includes, as random information, a character string included in the target registration information (i.e., a randomly generated character string).

[0111] The voice acquiring unit 250 acquires a response voice in response to the output information. For example, the voice acquiring unit 250 acquires a registration response voice that is a response voice in response to the registration output information. For example, the voice acquiring unit 250 acquires an authentication response voice that is a response voice in response to the authentication output information.

[0112] The authentication-related processing unit 270 performs authentication-related processing using at least one of voiceprint information and word information included in the response voice. For example, the authentication-related processing unit 270 registers the target using the voiceprint information and word information included in the registration response voice. For example, the authentication-related processing unit 270 authenticates the target using the voiceprint information and character string included in the authentication response voice and the registration information, respectively.

[0113] In detail, for example, authentication-related processing unit 270 includes registration unit 271 that performs registration processing and authentication unit 272 that performs authentication processing. An example will be described in which authentication-related processing unit 270 includes a voiceprint estimation model that estimates voiceprint information from a registration response voice.

[0114] The registration unit 271 registers the target using voiceprint information and word information included in the registration response voice. As shown in Fig. 9, the registration unit 271 includes a voiceprint acquisition unit 271a, a word acquisition unit 271b, and a registration control unit 271c, for example.

[0115] Voiceprint acquisition unit 271a acquires voiceprint information from the registration response voice using, for example, a voiceprint estimation model.

[0116] The word acquiring unit 271b acquires word information from the registration response voice. The word acquiring unit 271b may acquire word information from the registration response voice, for example, by using a word estimation model. The registration response voice may be a response voice in response to registration output information that includes a randomly generated character string. Therefore, the word information may include a character string included in the response voice in response to random information (for example, the registration response voice).

[0117] The registration control unit 271c generates registration information that associates the acquired voiceprint information and word information (character strings in this embodiment), and stores the registration information in the registration information storage unit 295. The registration control unit 271c may generate registration information that further associates the acquired target information, and store the registration information in the registration information storage unit 295. The word information used by the registration control unit 271c may be word information acquired by the word acquisition unit 271b, or may be word information acquired by the random information acquisition unit 210.

[0118] The authentication unit 272 authenticates the target using voiceprint information and word information (character strings in this embodiment) included in the authentication response voice and the registration information. The authentication unit 272 includes, for example, as shown in Fig. 10, a target identification unit 272a and a target authentication unit 272b.

[0119] The object specifying unit 272a specifies the object using the acquired object information and object information included in the registration information.

[0120] The target authentication unit 272b authenticates the identified target using voiceprint information and character strings included in the authentication response voice and the registration information of the identified target. The target authentication unit 272b includes, for example, a first authentication unit 272b1, a second authentication unit 272b2, and an authentication result output unit 272b3.

[0121] The first authentication unit 272b1 performs the first authentication of the target using voiceprint information included in the authentication response voice and the registration information of the identified target. For example, the first authentication unit 272b1 acquires voiceprint information from the authentication response voice using a voiceprint estimation model, and performs the first authentication of the target using the acquired voiceprint information and the voiceprint information included in the registration information.

[0122] The second authentication unit 272b2 performs second authentication of the target using character strings included in the authentication response voice and the registration information of the identified target. The authentication response voice used here may be, for example, a response voice to authentication output information including a randomly generated character string. Therefore, the word information may include a character string included in the response voice to the random information (for example, the authentication response voice).

[0123] The authentication result output unit 272b3 outputs the authentication result of the target using the results of the first authentication and second authentication of the target.

[0124] (Example of processing operation of information processing device 200) The information processing device 200 executes information processing including a first registration process and a first authentication process as shown in, for example, FIGS. 11 and 12, respectively.

[0125] (Example of the first registration process) 11 is a flowchart showing an example of a first registration process for registering a target. The registration process may be started, for example, by the target using terminal 50 to execute predetermined application software, inputting target information into a predetermined input field, and receiving an instruction to transmit the target information. Note that the trigger for starting the registration process is not limited to the example given here.

[0126] The terminal communication unit 59 transmits the target information received by the input unit 57 from the target who is the user of the terminal 50 to the information processing device 200 (step S59).

[0127] The target information acquisition unit 290 acquires the target information transmitted in step S59 (step S290).

[0128] The random information acquisition unit 210 acquires random information including a randomly generated character string (step S210).

[0129] For example, when the random information acquisition unit 210 receives the target information, it acquires a random character string used to register the target. Then, the random information acquisition unit 210 generates random information including the randomly generated character string.

[0130] The output unit 230 outputs output information (output information for registration) including the random information generated in step S210 (step S230a).

[0131] For example, the output unit 230 transmits the output information for registration to the terminal 50. The output information for registration includes a randomly generated character string as random information.

[0132] The output information for registration may be, for example, information that causes a randomly generated character string to be displayed on the terminal 50. Note that the output information for registration is not limited to information that causes a character string to be displayed, and may be, for example, information that causes a sound to be output from a speaker or the like of the terminal 50.

[0133] The display control unit 53 causes the display unit 55 to display the random information (step S53).

[0134] For example, when the display control unit 53 acquires the output information transmitted in step S230a from the output unit 230, it causes the display unit 55 to display the random information included in the output information.

[0135] In more detail, for example, assume that the random character string included in the random information is "j984jv0wk." The display control unit 53 displays, for example, "j984jv0wk" on the display unit 55. The output information may further include a message requesting that the character string be read. In this case, the display control unit 53 may display, for example, a message such as "Please read this character string aloud" along with the random character string "j984jv0wk."

[0136] The communication unit 51 transmits a response voice (a registration response voice) in response to the random information displayed in step S53 (step S51a).

[0137] For example, when the target reads the character string displayed in step S53, the communication unit 51 acquires the voice of the person reading the character string, and then transmits the acquired voice to the information processing device 200 as a response voice for registration.

[0138] The voice acquisition unit 250 acquires the response voice (registration response voice) transmitted in step S51a (step S250a).

[0139] The registration unit 271 registers the target using the voiceprint information and word information included in the response voice (response voice for registration) acquired in step S250a (step S271).

[0140] In step S271, for example, voiceprint acquiring unit 271a acquires voiceprint information from the registration response voice acquired in step S250a (step S271a).

[0141] The word acquiring unit 271b acquires word information from the registration response voice acquired in step S250a (step S271b), for example. As a result, word information including the random character string displayed on the display unit 55 is acquired from the registration response voice that reads the random character string.

[0142] The registration control unit 271c registers registration information that associates the object information, voiceprint information, and word information acquired in steps S290, S271a, and S271b (step S271c), for example. As a result, the registration control unit 271c ends step S271 and ends the registration process.

[0143] In detail, for example, the registration control unit 271c generates registration information that associates object information, voiceprint information, and word information. The registration control unit 271c may assign an object ID (identification), which is information for identifying the object, to the object, and generate registration information that further associates the object ID. Then, the registration control unit 271c stores the generated registration information in the registration information storage unit 295.

[0144] (Example of the first authentication process) 12 is a flowchart showing an example of a first authentication process for authenticating a target. The authentication process may be started, for example, when the target uses the terminal 50 to execute predetermined application software, inputs target information into a predetermined input field, and receives an instruction to transmit the target information.

[0145] The trigger for starting the authentication process is not limited to the example given here. For example, the authentication process may be executed to verify the identity of the user when talking with an operator face-to-face or over the phone (including a call using the call function of the terminal 50). When the authentication process is performed face-to-face, the terminal 50 may be, for example, a terminal used by the operator.

[0146] Steps S59 and S290 are executed in the same manner as in the first registration process.

[0147] The object specifying unit 272a specifies an object using the object information acquired in S290 and the object information included in the registration information (step S272a).

[0148] For example, the target identification unit 272a identifies, from the registration information stored in the registration information storage unit 295, registration information that includes target information that matches the target information acquired in S290 as the target registration information. Here, if the target registration information cannot be identified, the target identification unit 272a may determine that the authentication has failed and terminate the authentication process. Furthermore, if the target registration information can be identified, the next process (step S230b) may be executed.

[0149] The output unit 230 uses word information included in the target registration information identified in step S272a to output output information (authentication output information) including random information (step S230b).

[0150] For example, the output unit 230 transmits the output information to the terminal 50. The word information includes, for example, a randomly generated character string as described above. The output unit 230 outputs output information (authentication output information) that includes this character string in the random information. In other words, the authentication output information includes, as random information, a character string associated with the target information in the registration information of the identified target.

[0151] Similar to the above-described output information for registration, the output information for authentication may be, for example, information that displays a character string on the terminal 50. Note that the output information for authentication is not limited to information that displays a character string, and may be, for example, information that is output as sound from a speaker or the like of the terminal 50.

[0152] Step S53 is executed, which is the same as in the registration process. In addition, steps S51b and S250b are executed. Steps S51b and S250b correspond to the process in which "output information for registration" and "response voice for registration" in steps S51a and S250a of the first registration process are replaced with "output information for authentication" and "response voice for authentication," respectively.

[0153] The object authentication unit 272b authenticates the identified object using the voiceprint information and character string contained in the response voice (authentication response voice) acquired in step S250b and the registration information of the object identified in step S272a (step S272b).

[0154] In detail, for example, the first authentication unit 272b1 performs first authentication of the target using voiceprint information contained in the authentication response voice acquired in step S250b and the registration information of the target identified in step S272a (step S272b1).

[0155] For example, the first authentication unit 272b1 may perform the first authentication of the target using a similarity between the voiceprint information included in each of the authentication response voice and the registration information of the identified target and a predetermined first threshold. In detail, for example, the similarity may be at least one of cosine similarity, Mahalanobis distance, Euclidean distance, etc. Then, the first authentication unit 272b1 may determine that the first authentication of the target is successful if the similarity is equal to or greater than the first threshold, and may determine that the first authentication of the target is unsuccessful if the similarity is less than the first threshold.

[0156] The second authentication unit 272b2 performs second authentication of the target using character strings included in the authentication response voice and the registration information of the identified target (step S272b2).

[0157] For example, the second authentication unit 272b2 may perform the second authentication of the target based on whether or not a character string included in the authentication response voice matches a character string included in the registration information of the identified target. Here, the random information (character string) included in the authentication output information output in step S230b may be used as the character string included in the registration information.

[0158] In detail, for example, the second authentication unit 272b2 may determine that the second authentication of the target has been successful if these character strings match, and may determine that the second authentication of the target has failed if these character strings do not match.

[0159] The authentication result output unit 272b3 outputs the authentication result of the target using the results of the first authentication and second authentication of the target (step S272b3). As a result, the authentication result output unit 272b3 ends step S272b and ends the authentication process.

[0160] For example, the authentication result output unit 272b3 may determine that the authentication of the target is successful when both the first authentication and the second authentication of the target are successful. The authentication result output unit 272b3 may determine that the authentication of the target is unsuccessful when at least one of the first authentication and the second authentication of the target is unsuccessful.

[0161] Then, the authentication result output unit 272b3 may transmit an authentication result, for example, including whether authentication of the target was successful or unsuccessful, to the terminal 50. Upon acquiring the authentication result, the display control unit 53 may display the acquired authentication result on the display unit 55. This allows the target to know the authentication result.

[0162] (Actions and Effects) As described above, according to this embodiment, the random information includes a randomly generated character string. The word information includes a character string included in a response voice to the random information. The registration information is information that associates the target's voiceprint information with the randomly generated character string. The authentication-related processing unit 270 authenticates the target using the voiceprint information and character string included in the authentication response voice and the registration information, respectively.

[0163] This allows two-factor authentication to be performed using an authentication response voice for a randomly generated character string.

[0164] A random character string is more difficult for a third party to know than a fixed character string, making it more difficult for a third party to impersonate you.

[0165] Furthermore, when attempting to spoof a target's voice using AI to create a voice that mimics the target's voice, AI generally has difficulty processing meaningless phrases. Therefore, using AI to create a voice that reads meaningless phrases, such as random character strings, generally results in unnatural speech. Therefore, there is often a large difference (low similarity) in the voiceprint information obtained from an authentication response voice in which the target reads a random character string and an authentication response voice created by AI to imitate the voice of the target reading a random character string. Therefore, using an authentication response voice that is based on a randomly generated character string makes spoofing using AI more difficult. This enables robust authentication using the target's voice.

[0166] According to this embodiment, the information processing device 200 includes a target information acquisition unit 290 that acquires target information for identifying a target. The registration information is information that further associates the target information of the target. The authentication output information includes, as random information, a character string associated with the target information of the target.

[0167] The authentication unit 272 includes an object identification unit 272a and an object authentication unit 272b. The object identification unit 272a identifies an object using the acquired object information and object information included in the registration information. The object authentication unit 272b authenticates the identified object using voiceprint information and character strings included in the authentication response voice and the registration information of the identified object.

[0168] This allows for multi-step authentication, including a first step using target information and a second step using voiceprint information and word information (character strings). Furthermore, in the second step, two-factor authentication can be performed using a random character string used in the target registration process and voiceprint information obtained from the corresponding response voice. Therefore, stronger authentication using the target's voice becomes possible.

[0169] The second embodiment may be modified as follows.

[0170] (Variation 1) The registration control unit 271c may use the random information acquired in step S210 as the word information to be included in the registration information, instead of the word information acquired in step S271b. This also achieves the same effects as in the second embodiment.

[0171] (Variation 2) Before generating or registering the registration information, the registration control unit 271c may determine whether the word information (character string) acquired in step S271b matches the random information (character string) acquired in step S210.

[0172] Then, if the character strings contained in the word information and the random information match, the registration control unit 271c may, for example, generate registration information that associates the target information, voiceprint information, and word information, and store the information in the registration information storage unit 295. If the character strings contained in the word information and the random information do not match, the registration control unit 271c may repeatedly execute the processes of steps S210 to S271b described above until they match. An upper limit on the number of repetitions may be set in advance. If the character strings contained in the word information and the random information do not match even after repeating the process up to this upper limit, the registration control unit 271c may end the registration process without registering the target, i.e., without generating registration information for the target and storing it in the registration information storage unit 295.

[0173] This also provides the same effect as in the second embodiment.

[0174] (Variation 3) The registered information may be associated with a plurality of pieces of word information, each of which includes a randomly generated character string.

[0175] For example, the processes from steps S210 to S271 described above may be repeated N times, where N is an integer equal to or greater than 2. Then, the registration control unit 271c may generate registration information in which the target information is associated with multiple sets of voiceprint information and word information, and store the generated registration information in the registration information storage unit 295. The voiceprint information may be information acquired from multiple pieces of word information. In this case, the registration control unit 271c may generate registration information in which the target information, voiceprint information, and multiple pieces of word information are associated with each other, and store the generated registration information in the registration information storage unit 295.

[0176] Alternatively, for example, in step S210, a plurality of pieces of word information each including a randomly generated character string may be acquired, and in step S230a, output information including the plurality of pieces of word information may be output. This output information may further include, for example, a message requesting the subject to read a random character string included in each of the plurality of pieces of word information. Then, in steps S53 and S55, a registration response voice in which the subject reads the character string displayed on terminal 50 may be transmitted, and in step S250a, the registration response voice may be acquired.

[0177] In steps S271a and S271b, voiceprint information and word information for each of a plurality of pieces of word information (random character strings) may be acquired. In step S271c, the registration control unit 271c may generate registration information in which the target information is associated with a plurality of sets of voiceprint information and word information, and store the registration information in the registration information storage unit 295.

[0178] Alternatively, the voiceprint information may be information acquired from a plurality of pieces of word information. For example, in step S271a, voiceprint information for a plurality of pieces of word information (random character strings) may be acquired, and in step S271b, word information for each of the plurality of pieces of word information (random character strings) may be acquired. In this case, the registration control unit 271c may generate registration information that associates the target information, the voiceprint information, and the plurality of pieces of word information, and store the registration information in the registration information storage unit 295.

[0179] In the authentication process in this modified example, the target may be authenticated using at least one of a plurality of pieces of word information included in the registered information of the target identified in step S272a.

[0180] For example, in step S230b, the output unit 230 may select at least one word information from multiple word information included in the target registration information identified in step S272a, and output output information (authentication output information) including the selected word information as random information.

[0181] Then, in steps S53 and S55, an authentication response voice in which the subject reads the character string displayed on the terminal 50 may be transmitted, and in step S250a, the authentication response voice may be acquired.

[0182] In step S272b, the object authentication unit 272b may authenticate the identified object, for example, by using the voiceprint information and character string contained in the response voice (authentication response voice) acquired in step S250b and the registration information of the object identified in step S272a.

[0183] In detail, for example, it is assumed that the output information includes a plurality of pieces of word information (random character strings).

[0184] If registration information including multiple sets of voiceprint information and word information is stored, in step S272b, steps S272b1 and S272b2 may be executed for each of the multiple sets of word information.

[0185] If registration information including voiceprint information and information on a plurality of words is stored, in step S272b1, the first authentication unit 272b1 may acquire voiceprint information from at least one character string included in the authentication response voice. Then, the first authentication unit 272b1 may perform first authentication of the target using the acquired voiceprint information and the voiceprint information included in the registration information.

[0186] In step S272b2, the second authentication unit 272b2 may perform the second authentication of the target based on whether all of at least one character string included in the authentication response voice is included in the target's registration information. The second authentication unit 272b2 may determine that the second authentication of the target is successful if all of the character strings included in the authentication response voice are included in the target's registration information, and may determine that the second authentication of the target is unsuccessful if the character strings included in the authentication response voice include at least one unregistered character string. An unregistered character string is a character string that is not included in the target's registration information.

[0187] In step S272b3, for example, as described in embodiment 2, the authentication result output unit 272b3 may output the authentication result of the target using the results of the first authentication and second authentication of the target. Here, if at least one of the first authentication and the second authentication is performed multiple times, the authentication result output unit 272b3 may determine that the authentication of the target is successful if all of the first authentications and all of the second authentications are successful. The authentication result output unit 272b3 may determine that the authentication of the target is unsuccessful if at least one of the first authentications fails or at least one of the second authentications fails.

[0188] According to this modification, a target can be authenticated using a plurality of pieces of random information (random character strings), thereby enabling even stronger authentication using the target's voice.

[0189] [Embodiment 3] In the third embodiment, an example will be described in which the random information includes a randomly generated character string, similar to the second embodiment. Similarly to the second embodiment, the word information in the third embodiment also includes a character string included in a response voice in response to the random information, that is, a randomly generated character string.

[0190] This embodiment differs from embodiment 2 in that target information is acquired by voice (response voice). In this embodiment, an example is described in which voiceprint authentication is performed using voiceprint information acquired from the target information and response voices to random information, but voiceprint information may be acquired using either the response voice to target information or the response voice to random information. In this embodiment, for simplicity, explanations that overlap with other embodiments will be omitted as appropriate.

[0191] (Configuration example of information processing device 300) The information processing system SYS may include an information processing device 300 instead of the information processing device 100. The information processing device 300 includes, for example, a target information acquisition unit 390, a registration information storage unit 295 and a random information acquisition unit 210 similar to those in the second embodiment, an output unit 330, a voice acquisition unit 350, and an authentication-related processing unit 370, as shown in FIG.

[0192] The target information acquisition unit 390 acquires target information for identifying a target. The target information acquisition unit 390 acquires the target information by using, for example, a response voice in response to output information including a request for target information, which will be described later.

[0193] In detail, for example, the above-mentioned word estimation model may be a machine learning model that has been trained to obtain word information and target information from speech. The target information acquisition unit 390 may use the word estimation model to acquire target information from response speech in response to output information including a request for target information, which will be described later. Note that the method by which the target information acquisition unit 390 acquires target information from response speech is not limited to a method using a machine learning model such as a word estimation model, and general techniques may be used.

[0194] The output unit 330 outputs output information, which includes, for example, registration output information and authentication output information.

[0195] For example, the output information for registration includes a request for target information as a request for predetermined information, and also includes, for example, a randomly generated character string as random information.

[0196] For example, the authentication output information includes a request for target information as a request for predetermined information, and also includes, for example, a character string included in the target registration information (i.e., a randomly generated character string) as random information.

[0197] The voice acquisition unit 350 acquires a response voice in response to the output information. For example, the voice acquisition unit 350 acquires a registration response voice that is a response voice in response to the registration output information. For example, the voice acquisition unit 350 acquires an authentication response voice that is a response voice in response to the authentication output information.

[0198] For example, each of the registration response voice and the authentication response voice may include a response voice to registration output information that includes a request for target information, which is a request for predetermined information. For example, each of the registration response voice and the authentication response voice may include a response voice to registration output information that includes a character string that is random information.

[0199] The authentication-related processing unit 370 performs authentication-related processing using at least one of voiceprint information and word information included in the response voice. For example, the authentication-related processing unit 370 registers the target using the voiceprint information and word information included in the registration response voice. For example, the authentication-related processing unit 370 authenticates the target using the voiceprint information and character string included in the authentication response voice and the registration information, respectively.

[0200] In more detail, for example, authentication-related processing unit 370 includes registration unit 371 that performs registration processing and authentication unit 372 that performs authentication processing. An example will be described in which authentication-related processing unit 370 includes a voiceprint estimation model that estimates voiceprint information from a registration response voice.

[0201] The registration unit 371 registers the target using the voiceprint information and word information included in the registration response voice. As shown in Fig. 14, the registration unit 371 includes a voiceprint acquisition unit 371a, a word acquisition unit 271b, and a registration control unit 271c similar to those in the second embodiment.

[0202] The voiceprint acquiring unit 371a acquires voiceprint information from the registration response voice using, for example, a voiceprint estimation model. The voiceprint acquiring unit 371a may acquire voiceprint information from, for example, the registration response voice in response to the registration output information including a request for target information, and the registration response voice in response to the registration output information including a randomly generated character string.

[0203] In addition, the voiceprint acquisition unit 371a may acquire voiceprint information from either a registration response voice in response to registration output information that includes a request for target information, or a registration response voice in response to registration output information that includes a randomly generated character string.

[0204] The authentication unit 372 authenticates the target using the voiceprint information and character string included in the authentication response voice and the registration information. The authentication unit 372 includes, as shown in Fig. 15, a target identification unit 272a and a target authentication unit 372b similar to those in the second embodiment.

[0205] The target authentication unit 372b authenticates the identified target using voiceprint information and character strings included in the authentication response voice and the registration information of the identified target. The target authentication unit 372b includes, for example, a first authentication unit 372b1, a second authentication unit 272b2, and an authentication result output unit 272b3 similar to those in the second embodiment.

[0206] The first authentication unit 372b1 performs the first authentication of the target using voiceprint information included in the authentication response voice and the registration information of the identified target. For example, the first authentication unit 372b1 acquires voiceprint information from the authentication response voice using a voiceprint estimation model, and performs the first authentication of the target using the acquired voiceprint information and the voiceprint information included in the registration information.

[0207] The first authentication unit 372b1 may acquire voiceprint information from, for example, an authentication response voice in response to authentication output information including a request for target information and an authentication response voice in response to authentication output information including a randomly generated character string. That is, the first authentication unit 372b1 may use voiceprint information acquired from the authentication response voice in response to the request for target information and random information for authentication using voiceprint information (first authentication).

[0208] In addition, the first authentication unit 372b1 may acquire voiceprint information from either an authentication response voice in response to authentication output information including a request for target information, or an authentication response voice in response to authentication output information including a randomly generated character string.

[0209] (Example of processing operation of information processing device 300) The information processing device 300 executes information processing including a second registration process and a second authentication process as shown in, for example, FIGS. 16 and 17, respectively.

[0210] (Example of the second registration process) FIG. 16 is a flowchart showing an example of a second registration process for registering a target. The registration process may be initiated, for example, by the target using terminal 50 to execute predetermined application software, and terminal 50 receiving a request for target information and acquiring a response voice (registration response voice) in response to the request from the target. Here, upon receiving the request for target information, terminal 50 may provide the target with a message, such as "Please state your address and name," by voice or display, requesting the target to speak the target information. Note that the trigger for starting the registration process is not limited to the example given here.

[0211] The call unit 51 transmits a response voice including the target information, which is uttered by the target who is the user of the terminal 50, to the information processing device 200 (step S59a).

[0212] For example, when the target speaks target information corresponding to a request such as an address or a name, the communication unit 51 acquires the voice of the target information. Then, the communication unit 51 transmits the acquired voice to the information processing device 300 as a response voice for registration in response to the output information for registration including the request for the target information.

[0213] The voice acquisition unit 350 acquires the response voice transmitted in step S59b (step S310a).

[0214] For example, the voice acquisition unit 350 acquires a response voice (registration response voice) in response to the registration output information including a request for target information.

[0215] The target information acquisition unit 390 acquires target information from the response voice acquired in step S310a (step S390a).

[0216] For example, the target information acquisition unit 390 acquires the target information from a response voice for registration in response to output information for registration that includes a request for the target information.

[0217] Step S210 is executed in the same manner as in embodiment 2. The output unit 330 executes step S230a in the same manner as in embodiment 2. Steps S53 and S51a are executed in the same manner as in embodiment 2. The voice acquisition unit 350 executes step S250a in the same manner as in embodiment 2.

[0218] The registration unit 371 registers the target using the voiceprint information and word information included in the response voice (registration response voice) acquired in steps S310a and S250a (step S371).

[0219] In step S371, for example, the voiceprint acquiring unit 371a acquires voiceprint information from the registration response voice acquired in steps S310a and S250a (step S371a).

[0220] Step S271b is executed in the same manner as in the second embodiment.

[0221] The registration control unit 271c registers registration information that associates the object information, voiceprint information, and word information acquired in steps S390, S371a, and S271b (step S371c), for example. As a result, the registration control unit 271c ends step S371 and ends the registration process.

[0222] That is, in step S371c, the information acquired in steps S390 and S371a is used instead of steps S290 and S271a to register the registration information. Except for this, step S371c may be the same as step S271c in the second embodiment.

[0223] (Example of second authentication process) 17 is a flowchart showing an example of a second authentication process for authenticating a target. The authentication process may be initiated, for example, by the target using terminal 50 to execute predetermined application software, and terminal 50 receiving a request for target information and acquiring a response voice (authentication response voice) from the target in response to the request. Here, upon receiving the request for target information, terminal 50 may provide the target with a message, such as "Please state your address and name," by voice or display, requesting the target to speak the target information. Note that the trigger for starting the authentication process is not limited to the example given here.

[0224] Steps S59b, S310b, and S390b are executed. Steps S59b, S310b, and S390b correspond to the process in which the "registration output information" and "registration response voice" in steps S59a, S310a, and S390b of the second registration process are replaced with "authentication output information" and "authentication response voice," respectively.

[0225] Step S272a is executed in the same manner as in embodiment 2. The output unit 330 executes step S230b in the same manner as in embodiment 2. Steps S53 and S51b are executed in the same manner as in embodiment 2. The voice acquisition unit 350 executes step S250b in the same manner as in embodiment 2.

[0226] The target authentication unit 372b authenticates the identified target using the voiceprint information and word information contained in the response voice (authentication response voice) acquired in steps S310b and S250b and the registration information of the target identified in step S272a (step S372b).

[0227] In detail, for example, the first authentication unit 372b1 performs first authentication of the target using the voiceprint information contained in the authentication response voice acquired in steps S310 and S250b and the registration information of the target identified in step S272b (step S372b1).

[0228] For example, in step S372b1, "the authentication response voice acquired in steps S310 and S250b" is used instead of "the authentication response voice acquired in step S250b" in step S272b1 of embodiment 2. Except for this point, step S372b1 may be the same as step S272b1 of embodiment 2.

[0229] Steps S272b2 and S272b3 are executed in the same manner as in embodiment 2. As a result, the authentication result output unit 272b3 ends step S272b and ends the authentication process.

[0230] (Actions and Effects) As described above, according to this embodiment, the authentication output information includes a request for target information as a request for predetermined information, and includes, as random information, a character string associated with the target information of the identified target. The target information acquisition unit 390 acquires the target information using an authentication response voice in response to the request for target information. The target authentication unit 372b uses voiceprint information acquired from the authentication response voice in response to the request for target information and the random information for authentication using voiceprint information.

[0231] This allows voiceprint authentication to be performed using voiceprint information obtained from a request for target information and an authentication response voice in response to random information. Therefore, voiceprint information can be obtained from a longer response voice than when only an authentication response voice in response to random information is used, allowing for more robust authentication using the target voice.

[0232] [Embodiment 4] In the fourth embodiment, an example will be described in which the random information includes a second question randomly generated in relation to the first question. In the present embodiment, an example will be described in which the word information includes a keyword included in a response voice (answer) to at least one of the first question and the second question. In the present embodiment, for the sake of brevity, explanations that overlap with other embodiments will be omitted as appropriate.

[0233] (Configuration example of information processing device 400) The information processing system SYS may include an information processing device 400 instead of the information processing device 100. The information processing device 400 includes, for example, as shown in Fig. 18 , a target information acquisition unit 290, a registration information storage unit 495, a random information acquisition unit 410, an output unit 430, a voice acquisition unit 450, and an authentication-related processing unit 470, similar to those in the second embodiment.

[0234] The registration information storage unit 495 is a storage unit for storing registration information of a target. The registration information may be information that associates the voiceprint information of the target with a question and a keyword for the question. The question may include a first question. In this case, the keyword may include a phrase included in the response voice (answer) to the first question. The question may also include a second question. In this case, the keyword may include a phrase included in the response voice (answer) to the second question. The registration information may be information that further associates the target information of the target.

[0235] The random information acquisition unit 410 acquires random information including a randomly generated second question.

[0236] The output unit 430 outputs output information. The output information includes, for example, registration output information and authentication output information. For example, the registration output information includes a first question. For example, the authentication output information includes the first question, which is a question registered in the registration information, and a second question, which is randomly generated in relation to the first question, as random information. Note that the first question registered in the registration information may be randomly generated information, as described above.

[0237] The voice acquisition unit 450 acquires a response voice in response to the output information. For example, the voice acquisition unit 450 acquires a registration response voice, which is a response voice in response to the registration output information. The registration response voice includes, for example, an answer to the first question. For example, the voice acquisition unit 450 acquires an authentication response voice, which is a response voice in response to the authentication output information. The authentication response voice includes, for example, an answer to the first question and an answer to the second question.

[0238] The authentication-related processing unit 470 performs authentication-related processing using at least one of voiceprint information and word information contained in the response voice.

[0239] For example, the authentication-related processing unit 470 registers the target using voiceprint information included in the response voice for registration and a keyword in response to the first question. This keyword may be, for example, a keyword included in the answer to the first question.

[0240] For example, authentication-related processing unit 470 authenticates the target using voiceprint information included in the authentication response voice and the registration information, and a keyword for the first question. This keyword may be, for example, a keyword included in the answer to the first question. For example, authentication-related processing unit 270 may register a second question and a keyword for the second question as the registration information of the target. This keyword may be, for example, a keyword included in the answer to the second question.

[0241] In detail, for example, authentication-related processing unit 470 includes registration unit 471 that performs registration processing and authentication unit 472 that performs authentication processing. An example will be described in which authentication-related processing unit 470 includes a voiceprint estimation model that estimates voiceprint information from a registration response voice.

[0242] The registration unit 471 registers the target using voiceprint information and word information (keywords) included in the registration response voice. The registration unit 471 includes a voiceprint acquisition unit 471a, a word acquisition unit 471b, and a registration control unit 471c, as shown in Fig. 19, for example.

[0243] Voiceprint acquiring unit 471a acquires voiceprint information from the registration response voice using, for example, a voiceprint estimation model.

[0244] The word acquiring unit 471b acquires word information from the registration response voice. The word acquiring unit 471b may acquire the word information from the registration response voice by using, for example, a word estimation model. This registration response voice may be a response voice (answer) to the registration output information including a question. The word information may include a keyword included in the response voice (answer) to the registration output information including a question. The question may include, for example, at least one of a first question and a second question. That is, the word information may include, for example, a keyword for the first question. The word information may include, for example, a keyword for the second question.

[0245] The registration control unit 471c generates registration information that associates the acquired voiceprint information and word information (keywords in this embodiment), and stores the information in the registration information storage unit 495. The word information used by the registration control unit 471c may be, for example, the word information acquired by the word acquisition unit 471b. The registration control unit 271c may generate registration information that further associates a question included in the output information, and store the information in the registration information storage unit 295. The registration control unit 271c may generate registration information that further associates the acquired target information, and store the information in the registration information storage unit 295.

[0246] The authentication unit 472 authenticates the target using voiceprint information and word information (keywords in this embodiment) included in the authentication response voice and the registration information. The authentication unit 472 includes, as shown in Fig. 20, a target identification unit 272a, a target authentication unit 472b, and a registration update unit 472c, which are the same as those in the second embodiment.

[0247] The target authentication unit 472b authenticates the identified target using voiceprint information included in the authentication response voice and the registration information of the identified target, and a keyword in response to the first question. The target authentication unit 472b includes, for example, a first authentication unit 472b1, a second authentication unit 472b2, and an authentication result output unit 272b3 similar to that of the second embodiment.

[0248] The first authentication unit 472b1 performs the first authentication of the target using voiceprint information included in the authentication response voice and the registration information of the identified target. For example, the first authentication unit 472b1 acquires voiceprint information from the authentication response voice using a voiceprint estimation model, and performs the first authentication of the target using the acquired voiceprint information and the voiceprint information included in the registration information.

[0249] The second authentication unit 472b2 performs the second authentication of the target using keywords included in the authentication response voice and the registration information of the identified target. The authentication response voice used here may be, for example, a response voice to authentication output information including a first question. In this case, the keyword may be, for example, a keyword included in the authentication response voice (answer) to the first question.

[0250] The second authentication unit 472b2 may perform the second authentication of the target using a keyword included in the authentication response voice (answer) to the authentication output information including a randomly generated question. Such second authentication may be, for example, the second authentication when the first question is a randomly generated question, or may be the second authentication after a randomly generated question (for example, the second question) and a keyword corresponding to the question are registered in the registration information. Therefore, the keyword used for the second authentication may be a keyword included in the response voice to the random information, that is, for example, the authentication response voice (answer) to the randomly generated question.

[0251] The registration update unit 472c registers the second question and the keyword for the second question as the target registration information.

[0252] (Example of processing operation of information processing device 400) The information processing device 400 executes information processing including a third registration process and a third authentication process as shown in, for example, FIGS. 21 and 22, respectively.

[0253] (Example of the third registration process) FIG. 21 is a flowchart illustrating an example of a third registration process for registering an object.

[0254] Steps S59 and S290 are executed in the same manner as in the second embodiment.

[0255] The output unit 430 outputs output information (output information for registration) including a question (for example, a first question) (step S430a).

[0256] Here, the first question may be a randomly generated question. In this case, the random information acquisition unit 410 may acquire the first question. Then, the output unit 430 may output output information (output information for registration) including the first question acquired by the random information acquisition unit 410.

[0257] The display control unit 53 causes the display unit 55 to display the question (step S453a).

[0258] For example, when the display control unit 53 acquires the output information transmitted in step S430a from the output unit 230, the display control unit 53 causes the display unit 55 to display the question included in the output information.

[0259] In detail, for example, the display control unit 53 causes the display unit 55 to display the first question included in the output information, "What was the name of the first pet you had?"

[0260] The question included in the output information may be provided to the target via, for example, the terminal 50, and the method for providing the question is not limited to displaying it. For example, the communication unit 51 may utter the first question to the target by voice, or the question may be output as voice from a speaker or the like of the terminal 50.

[0261] The communication unit 51 transmits a response voice (a registration response voice) to the question displayed in step S53 (step S451a).

[0262] For example, when the subject answers the first question displayed in step S453a by voice, the communication unit 51 acquires the voice including the answer. Then, the communication unit 51 transmits the acquired voice to the information processing device 400 as a registration response voice.

[0263] The voice acquiring unit 450 acquires the response voice (registration response voice) transmitted in step S451a (step S450a).

[0264] The registration unit 471 registers the target using the voiceprint information and word information included in the response voice (registration response voice) acquired in step S450a and the question included in the output information output in step S430a (step S471).

[0265] In step S471, for example, voiceprint acquiring unit 471a acquires voiceprint information from the registration response voice acquired in step S450a (step S471a).

[0266] The word acquiring unit 471b acquires word information from the response voice for registration acquired in step S450a (step S471b), for example. As a result, word information including a keyword is acquired from the response voice for registration (answer) to the question (for example, the first question) displayed on the display unit 55.

[0267] For example, suppose the subject answers "My first pet's name is Marie" to a first question such as "What was the name of the first pet you had?" or "Please tell me the name of the first pet you had." The word acquisition unit 471b acquires the keyword "Marie" from this answer using, for example, a word estimation model.

[0268] The registration control unit 471c registers registration information that associates the object information, voiceprint information, and word information acquired in steps S290, S471a, and S471b (step S471c), for example. As a result, the registration control unit 471c ends step S271 and ends the registration process.

[0269] In detail, for example, the registration control unit 471c generates registration information that associates the target information and voiceprint information acquired in steps S290 and S471a. The registration control unit 471c generates registration information that further associates the question (first question) included in the output information output in step S430a with the word information (i.e., keywords for the question) acquired in step S471b. The registration control unit 471c may assign a target ID to the target and generate registration information that further associates the target ID. Then, the registration control unit 471c stores the generated registration information in the registration information storage unit 495.

[0270] Note that voiceprint acquiring unit 471a may acquire voiceprint information and frequency information from the registration response voice acquired in step S450a. The frequency information is information indicating the frequency (pitch) of the response voice. A general technique may be used to acquire the frequency information from the response voice. In this case, the registration information may be information that further associates the frequency information of the voice (e.g., the registration response voice) in response to the first question.

[0271] It is generally known that the pitch of a voice changes depending on emotions. This makes it possible to register information that corresponds to the emotion of the subject in response to a question (first question). As will be described later, authentication can be performed using the emotion of the subject in response to the question (first question).

[0272] (Example of the third authentication process) FIG. 22 is a flowchart illustrating an example of a fourth authentication process for authenticating a target.

[0273] Steps S59, S290 and S272a are executed in the same manner as in the second embodiment.

[0274] The random information acquisition unit 410 acquires random information including a randomly generated question (for example, a second question) (step S410).

[0275] For example, the random information acquisition unit 210 acquires at least one of a question (e.g., a first question) and a corresponding keyword included in the registration information of the target identified in step S272a. For example, the random information acquisition unit 210 inputs a prompt using at least one of the first question and a keyword into the generation AI to acquire a randomly generated second question. The second question is different from the first question and is related to the first question. For example, the random information acquisition unit 210 generates random information including the second question.

[0276] For example, suppose the target registration information contains the keyword "Marie," which is the answer to the first question, "What was the name of your first pet?" In this case, the second question may be, for example, "What is Marie's favorite food?"

[0277] The output unit 430 outputs output information (authentication output information) including the random information acquired in step S410 (step S430b).

[0278] For example, output unit 430 transmits output information to terminal 50. The authentication output information includes, as random information, for example, a randomly generated second question. Furthermore, output unit 430 may cause random information acquisition unit 410 to identify the first question included in the target registration information identified in step S272a. Then, output unit 430 may output output information including the first question included in the target registration information and the randomly generated second question.

[0279] Similar to the above-described output information for registration, the output information for authentication may be, for example, information that displays a character string on the terminal 50. Note that the output information for authentication is not limited to information that displays a character string, and may be, for example, information that is output as sound from a speaker or the like of the terminal 50.

[0280] The display control unit 53 causes the display unit 55 to display the question (step S453b).

[0281] For example, when the display control unit 53 acquires the output information transmitted in step S430a from the output unit 230, the display control unit 53 causes the display unit 55 to display the question included in the output information.

[0282] In detail, for example, the display control unit 53 causes the display unit 55 to display a first question included in the output information, "What is the name of the first pet you had?". Also, for example, the display control unit 53 causes the display unit 55 to display a second question, "What is Marie's favorite food?". These questions may be displayed on one screen or may be displayed sequentially.

[0283] The question included in the output information may be provided to the target via, for example, the terminal 50, and the method for providing the question is not limited to displaying it. For example, the communication unit 51 may utter the first question to the target by voice, or the question may be output as voice from a speaker or the like of the terminal 50.

[0284] The communication unit 51 transmits a response voice (a registration response voice) to the question displayed in step S53 (step S451a).

[0285] For example, when the target responds to the first question and the second question displayed in step S453a by voice, such as "It's Marie" and "It's XX," respectively, the communication unit 51 acquires the voice including the response. Then, the communication unit 51 transmits the acquired voice to the information processing device 400 as a response voice for registration.

[0286] The voice acquiring unit 450 acquires the response voice (authentication response voice) transmitted in step S451a (step S450b).

[0287] The object authentication unit 472b authenticates the identified object using the voiceprint information and word information contained in the response voice (authentication response voice) acquired in step S450b and the registration information of the object identified in step S472a (step S472b).

[0288] In detail, for example, the first authentication unit 472b1 performs first authentication of the target using voiceprint information contained in the authentication response voice acquired in step S450b and the registration information of the target identified in step S472a (step S472b1).

[0289] For example, the first authentication unit 472b1 may perform the first authentication of the target by using a similarity between the voiceprint information included in the authentication response voice and the registration information of the identified target and a predetermined first threshold. In detail, for example, the first authentication unit 472b1 may determine whether the first authentication is successful or unsuccessful in the same manner as the above-described first authentication unit 272b1.

[0290] The second authentication unit 472b2 performs second authentication of the target by using keywords included in the authentication response voice and the registration information of the identified target (step S472b2).

[0291] For example, the second authentication unit 472b2 may perform the second authentication of the target based on whether or not keywords included in the authentication response voice (answer) to the first question and the registered information of the identified target match.

[0292] In detail, for example, the second authentication unit 272b2 may determine that the second authentication of the target has been successful if these keywords match, and may determine that the second authentication of the target has failed if these keywords do not match.

[0293] In the above example of the first question, it is assumed that "Marie" is registered in the registration information as a keyword for the first question. The second authentication unit 272b2 determines that the second authentication is successful, for example, when the keyword included in the authentication response voice (answer) for the first question matches "Marie" included in the target registration information.

[0294] The second authentication unit 472b2 may acquire a keyword and frequency information for the first question from the authentication response voice. Then, the second authentication unit 472b2 may perform the second authentication of the target using the keyword and frequency information for the first question included in the authentication response voice and the registration information. For example, the second authentication unit 472b2 may determine that the second authentication of the target is successful if the keywords match and the difference in frequency information is within a predetermined range. Alternatively, for example, the second authentication unit 472b2 may determine that the second authentication of the target is unsuccessful if the keywords do not match or the difference in frequency information exceeds a predetermined range. In this way, the authentication-related processing unit 470 may perform authentication of the target using the keyword and frequency information for the first question included in the authentication response voice and the registration information. This allows authentication to be performed using the target's emotion regarding the question (first question).

[0295] The authentication result output unit 472b3 outputs the authentication result of the target using the results of the first authentication and second authentication of the target (step S472b3).

[0296] The authentication result output unit 472b3 may determine whether the target authentication is successful or unsuccessful, and output the authentication result, for example, in the same manner as the above-described authentication result output unit 272b3.

[0297] The registration update unit 472c registers the second question and the keyword corresponding to the second question as the target registration information (Step S472c), thereby ending the authentication process.

[0298] For example, it is assumed that the registration update unit 472c acquires an authentication response voice (answer) of "It's XX" in response to the second question. The registration update unit 472c may, for example, input the answer "It's XX" into a word estimation model to acquire a keyword "XX". Then, the registration update unit 472c may associate the second question with the acquired keyword and store them in the registration information storage unit 495 as the target registration information.

[0299] In the next authentication process for the target, for example, the second question may be used instead of the first question in the above authentication process, or a random third question may be newly acquired instead of the above second question.

[0300] (Actions and Effects) As described above, according to this embodiment, the registration information is information that associates the voiceprint information of the target with a question and a keyword corresponding to the question. The authentication output information includes, as random information, a second question randomly generated in association with the first question, along with a first question that is a question registered in the registration information. The authentication-related processing unit 470 authenticates the target using the voiceprint information and the keyword corresponding to the first question contained in the authentication response voice and the registration information, respectively, and registers the second question and the keyword corresponding to the second question as the registration information of the target.

[0301] This allows two-factor authentication to be performed the next time using the authentication response audio for the randomly generated second question. Because the second question is random, it is difficult for a third party to know it, making it difficult for a third party to impersonate you.

[0302] Therefore, by using a response voice for authentication corresponding to a randomly generated character string, it becomes difficult for a third party to impersonate the user, thereby enabling strong authentication using the target voice.

[0303] Furthermore, since the authentication output information includes two questions, the first question and the second question, voiceprint information can be obtained from a longer response voice than when there is only one question, allowing for voiceprint authentication. This enables robust authentication using the target voice. Voiceprint information can be obtained from the long response voice and used for voiceprint authentication, enabling stronger authentication using the target's voice.

[0304] According to this embodiment, the information processing device 400 includes a target information acquisition unit 290 that acquires target information for identifying a target. The registration information is information that further associates the target information. The authentication-related processing unit 470 includes a target identification unit 272a, a target authentication unit 472b, and a registration update unit 472c. The target identification unit 272a identifies a target using the acquired target information and the target information included in the registration information. The target authentication unit 472b authenticates the identified target using voiceprint information included in the authentication response voice and the registration information of the identified target, and a keyword in response to the first question. The registration update unit 472c registers the second question and the keyword in response to the second question as the registration information of the target.

[0305] This allows for multi-step authentication, including a first step using target information and a second step using voiceprint information and word information (keywords). Furthermore, in the next authentication, two-factor authentication can be performed using an authentication voice response to a randomly generated second question. Therefore, stronger authentication using the target's voice becomes possible.

[0306] According to the present embodiment, the registration information is information that further associates frequency information of the voice in response to the first question. The authentication-related processing unit 470 performs authentication of the target by using the authentication response voice and the keyword and frequency information in response to the first question that are included in the registration information.

[0307] This allows authentication to be performed using the subject's emotion in response to the question (first question), thereby enabling more robust authentication using the subject's voice.

[0308] [Embodiment 5] In the fifth embodiment, an example will be described in which the random information includes a second question randomly generated in relation to the first question, as in the fourth embodiment. In the present embodiment, an example will be described in which the word information also includes keywords included in a response voice (answer) to at least one of the first question and the second question, as in the fourth embodiment.

[0309] This embodiment differs from embodiment 4 in that target information is acquired by voice (response voice). In this embodiment, an example is described in which voiceprint authentication is performed using voiceprint information acquired from the target information and response voices to random information, but voiceprint information may be acquired using either the response voice to target information or the response voice to random information. In this embodiment, to simplify the explanation, explanations that overlap with other embodiments will be omitted as appropriate.

[0310] (Configuration example of information processing device 500) The information processing system SYS may include an information processing device 500 instead of the information processing device 100. As shown in Fig. 23 , for example, the information processing device 500 includes a target information acquisition unit 390 similar to that of the third embodiment, a registration information storage unit 495 and a random information acquisition unit 410 similar to those of the fourth embodiment, an output unit 530, a voice acquisition unit 550, and an authentication-related processing unit 570.

[0311] The output unit 530 outputs output information, which includes, for example, registration output information and authentication output information.

[0312] For example, the output information for registration includes a request for target information as a request for predetermined information, and also includes, for example, a first question.

[0313] For example, the authentication output information includes a request for target information as a request for predetermined information. The authentication output information also includes, for example, a first question registered in the registration information and a second question randomly generated in association with the first question as random information. Note that the first question registered in the registration information may be random information generated randomly, as described above.

[0314] That is, the authentication output information may include a request for target information as a request for predetermined information, in addition to the first question and the second question.

[0315] The voice acquiring unit 550 acquires a response voice in response to the output information. For example, the voice acquiring unit 550 acquires a registration response voice that is a response voice in response to the registration output information. For example, the voice acquiring unit 550 acquires an authentication response voice that is a response voice in response to the authentication output information.

[0316] For example, each of the registration response voice and the authentication response voice may include a response voice to the registration output information including a request for target information, which is a request for predetermined information. The registration response voice may include, for example, an answer to a first question. The authentication response voice may include, for example, an answer to the first question and an answer to a second question.

[0317] The authentication-related processing unit 570 performs authentication-related processing using at least one of voiceprint information and word information contained in the response voice.

[0318] For example, the authentication-related processing unit 570 registers the target using voiceprint information included in the response voice for registration and a keyword in response to the first question. This keyword may be, for example, a keyword included in the answer to the first question.

[0319] For example, authentication-related processing unit 470 authenticates the target using voiceprint information included in the authentication response voice and the registration information, and a keyword for the first question. This keyword may be, for example, a keyword included in the answer to the first question. For example, authentication-related processing unit 270 may register a second question and a keyword for the second question as the registration information of the target. This keyword may be, for example, a keyword included in the answer to the second question.

[0320] In detail, for example, authentication-related processing unit 570 includes registration unit 571 that performs registration processing and authentication unit 572 that performs authentication processing. An example will be described in which authentication-related processing unit 570 includes a voiceprint estimation model that estimates voiceprint information from a registration response voice.

[0321] The registration unit 571 registers the target using voiceprint information and word information (keywords) included in the registration response voice. As shown in Fig. 24, the registration unit 471 includes a voiceprint acquisition unit 571a, a word acquisition unit 471b, and a registration control unit 471c similar to those in the fourth embodiment.

[0322] The voiceprint acquiring unit 571a acquires voiceprint information from the registration response voice using, for example, a voiceprint estimation model. The voiceprint acquiring unit 571a may acquire voiceprint information from, for example, the registration response voice in response to the registration output information including a request for target information and the registration response voice in response to the registration output information including a question (for example, the first question).

[0323] The voiceprint acquiring unit 371a may acquire voiceprint information from either a registration response voice in response to registration output information including a request for target information, or a registration response voice in response to registration output information including a question.

[0324] The authentication unit 572 authenticates the target using the voiceprint information and character string included in the authentication response voice and the registration information. As shown in Fig. 25, the authentication unit 572 includes a target identification unit 272a similar to that in the second embodiment, a target authentication unit 572b, and a registration control unit 471c similar to that in the fourth embodiment.

[0325] The target authentication unit 572b authenticates the identified target using voiceprint information included in the authentication response voice and the registration information of the identified target, and a keyword for the first question. The target authentication unit 572b includes, for example, a first authentication unit 572b1, a second authentication unit 472b2 similar to that in the fourth embodiment, and an authentication result output unit 272b3 similar to that in the second embodiment.

[0326] The first authentication unit 572b1 performs the first authentication of the target using voiceprint information included in the authentication response voice and the registration information of the identified target. For example, the first authentication unit 572b1 acquires voiceprint information from the authentication response voice using a voiceprint estimation model, and performs the first authentication of the target using the acquired voiceprint information and the voiceprint information included in the registration information.

[0327] The first authentication unit 572b1 may acquire voiceprint information from, for example, an authentication response voice in response to authentication output information including a request for target information and an authentication response voice in response to authentication output information including a first question and a randomly generated second question. That is, the first authentication unit 572b1 may use voiceprint information acquired from the authentication response voices in response to the request for target information, the first question, and the second question for authentication using voiceprint information (first authentication).

[0328] In addition, the first authentication unit 572b1 may acquire voiceprint information from an authentication response voice in response to at least one of the authentication output information including a request for target information, the authentication output information including a first question, and the authentication output information including a second question.

[0329] (Example of processing operation of information processing device 500) The information processing device 500 executes information processing including a fourth registration process and a fourth authentication process as shown in, for example, FIGS. 26 and 27, respectively.

[0330] FIG. 26 is a flowchart illustrating an example of a fourth registration process for registering an object.

[0331] Steps S59a, S310a, and S390a are executed in the same manner as in the third embodiment. Steps S430a, S453a, S451a, and S450a are executed in the same manner as in the fourth embodiment.

[0332] The registration unit 571 registers the target using the voiceprint information and word information contained in the response voice (registration response voice) acquired in steps S310a and S450a and the question contained in the output information output in step S430a (step S571).

[0333] In step S571, for example, voiceprint acquiring unit 571a acquires voiceprint information from the registration response voice acquired in steps S310a and S450a (step S571a).

[0334] Steps S471b and S471c are executed in the same manner as in embodiment 4. As a result, the registration control unit 471c ends step S271 and ends the registration process.

[0335] (Example of the fourth authentication process) FIG. 27 is a flowchart illustrating an example of a fourth authentication process for authenticating a target.

[0336] Steps S59b, S310b, S390b, and S272a are executed as in the third embodiment. Steps S410, S430b, S453b, S451b, and S450b are executed as in the fourth embodiment.

[0337] The target authentication unit 572b authenticates the identified target using the voiceprint information and word information contained in the response voice (authentication response voice) acquired in steps S310b and S450b and the registration information of the target identified in step S272a (step S572b).

[0338] In detail, for example, the first authentication unit 572b1 performs first authentication of the target using the voiceprint information contained in the authentication response voice acquired in steps S310b and S450b and the registration information of the target identified in step S272a (step S572b1).

[0339] For example, the first authentication unit 472b1 may perform the first authentication of the target by using a similarity between the voiceprint information included in the authentication response voice and the registration information of the identified target and a predetermined first threshold. In detail, for example, the first authentication unit 472b1 may determine whether the first authentication is successful or unsuccessful in the same manner as the above-described first authentication unit 272b1.

[0340] Steps S472b2 and S472b3 are executed in the same manner as in the fourth embodiment.

[0341] Step S472c is executed in the same manner as in embodiment 4. As a result, the registration update unit 472c ends the authentication process.

[0342] (Actions and Effects) As described above, according to this embodiment, the authentication output information includes a request for target information as a request for predetermined information, along with the first question and the second question. The target information acquisition unit 390 acquires the target information using an authentication response voice in response to the request for target information. The target authentication unit 572b uses voiceprint information acquired from the request for target information and the authentication response voice in response to the first and second questions for authentication using voiceprint information.

[0343] This allows voiceprint authentication to be performed using voiceprint information acquired from the authentication response voices to the request for target information, the first question, and the second question. Therefore, voiceprint information can be obtained from a response voice that is longer than the response voice for one or more of the request for target information, the first question, and the second question, and voiceprint authentication can be performed using the voice of the target. This enables more robust authentication using the target voice.

[0344] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0345] In addition, although the flowcharts used in the above description show multiple steps (processes) in a sequential order, the order of the steps executed in each embodiment is not limited to the order shown. In each embodiment, the order of the steps shown in the drawings can be changed as long as it does not cause any problems in terms of the content.

[0346] Some or all of the above-described embodiments can be described as, but are not limited to, the following supplementary notes. 1. an output means for outputting output information including randomly generated random information; a voice acquisition means for acquiring a response voice in response to the output information; and an authentication-related processing means for performing processing related to authentication of a target using at least one of voiceprint information and word information included in the response voice. Information processing device. 2. further comprising a random information acquisition means for acquiring the random information, the random information includes at least one of a second question randomly generated in relation to the first question and a randomly generated character string; The word information includes at least one of a keyword included in the response voice to the second question and the randomly generated character string. 1. An information processing device according to claim 1. 3. The output information further includes a predetermined request for information. 1. The information processing device according to 1. or 2. 4. The authentication-related processing means performs processing related to the authentication of the target using voiceprint information and word information included in the response voice. 1. The information processing device according to 1. or 2. 5. the output information includes registration output information that is output to register the target; the response voice includes a registration response voice that is a response voice to the registration output information, The process related to authenticating the target includes registering voiceprint information and word information included in the registration response voice as registration information of the target. 3. An information processing device according to the present invention. 6. the output information includes authentication output information that is output to authenticate the target; the response voice includes an authentication response voice that is a response voice to the authentication output information, The process related to authenticating the target includes authenticating the target using voiceprint information and word information included in the authentication response voice and the registration information. 5. An information processing device according to the present invention. 7. the random information includes a randomly generated string of characters; the word information includes a character string included in the response voice in response to the random information, the registration information is information that associates the voiceprint information of the subject with the randomly generated character string; The authentication-related processing means authenticates the target using voiceprint information and character strings included in the authentication response voice and the registration information. 6. An information processing device according to the present invention. 8. further comprising a target information acquisition means for acquiring target information for identifying the target, The registration information is information that further associates the target information of the target, the authentication output information includes, as the random information, the character string associated with the target information in the registration information of the target; The authentication-related processing means an object specifying means for specifying the object using the acquired object information and object information included in the registration information; and an object authentication means for authenticating the specified object by using voiceprint information and a character string included in the authentication response voice and the registration information of the specified object. 7. An information processing device according to the present invention. 9. the authentication output information includes, as the request for the predetermined information, a request for the target information, and includes, as the random information, the character string associated with the target information of the specified target; the target information acquisition means acquires the target information using the authentication response voice in response to the request for the target information, The target authentication means uses voiceprint information acquired from the authentication response voice in response to the request for target information and the random information for authentication using the voiceprint information. 8. An information processing device according to the present invention. 10. the registration information is information that associates the voiceprint information of the subject with a question and a keyword for the question; the authentication output information includes a first question that is a question registered in the registration information, and a second question that is randomly generated in association with the first question as the random information; The authentication-related processing means authenticates the target using voiceprint information and a keyword for a first question included in the authentication response voice and the registration information, and registers the second question and the keyword for the second question as registration information of the target. 6. An information processing device according to the present invention. 11. further comprising a target information acquisition means for acquiring target information for identifying the target, The registration information is information that further associates the target information, The authentication-related means includes: an object specifying means for specifying the object using the acquired object information and object information included in the registration information; an object authentication means for authenticating the specified object using voiceprint information included in the authentication response voice and the registration information of the specified object, and a keyword in response to the first question; and a registration update means for registering the second question and a keyword for the second question as the registration information of the target. 10. An information processing device according to the present invention. 12. the authentication output information includes, together with the first question and the second question, a request for the target information as the request for the predetermined information; the target information acquisition means acquires the target information using the authentication response voice in response to the request for the target information, The target authentication means uses voiceprint information acquired from the authentication response voices to the request for target information, the first question, and the second question for authentication using the voiceprint information. 10. An information processing device according to the present invention. 13. the registration information is information that further associates frequency information of the voice in response to the first question, The authentication-related processing means performs authentication of the target by using the authentication response voice and a keyword and frequency information for the first question included in the registration information. 10. An information processing device according to the present invention. 14. an information processing device that performs processing related to authentication of a target; a terminal used by the target for a call; The information processing device includes: an output means for outputting output information including randomly generated random information to the terminal; a voice acquisition means for acquiring the response voice in response to the output information from the terminal; an authentication-related processing means for performing a process related to the authentication of the target using at least one of voiceprint information and word information included in the response voice, The terminal a communication means for transmitting the response voice Information processing system. 15. The information processing device includes: further comprising a random information acquisition means for acquiring the random information, the random information includes at least one of a second question randomly generated in relation to the first question and a randomly generated character string; The word information includes at least one of a keyword included in the response voice to the second question and the randomly generated character string. 14. An information processing system according to the present invention. 16. The output information further includes a predetermined request for information. 14. An information processing system according to 15. 17. The authentication-related processing means performs processing related to the authentication of the target using voiceprint information and word information included in the response voice. 14. An information processing system according to 15. 18. the output information includes registration output information that is output to register the target; the response voice includes a registration response voice that is a response voice to the registration output information, The process related to authenticating the target includes registering voiceprint information and word information included in the registration response voice as registration information of the target. 16. An information processing system according to the present invention. 19. the output information includes authentication output information that is output to authenticate the target; the response voice includes an authentication response voice that is a response voice to the authentication output information, The process related to authenticating the target includes authenticating the target using voiceprint information and word information included in the authentication response voice and the registration information. 18. An information processing system according to the present invention. 20. the random information includes a randomly generated string of characters; the word information includes a character string included in the response voice in response to the random information, the registration information is information that associates the voiceprint information of the subject with the randomly generated character string; The authentication-related processing means authenticates the target using voiceprint information and character strings included in the authentication response voice and the registration information. 19. An information processing system according to the present invention. twenty one. The information processing device includes: further comprising a target information acquisition means for acquiring target information for identifying the target, The registration information is information that further associates the target information of the target, the authentication output information includes, as the random information, the character string associated with the target information in the registration information of the target; The authentication-related processing means an object specifying means for specifying the object using the acquired object information and object information included in the registration information; and an object authentication means for authenticating the specified object by using voiceprint information and a character string included in the authentication response voice and the registration information of the specified object. 20. An information processing system according to the present invention. twenty two. the authentication output information includes, as the request for the predetermined information, a request for the target information, and includes, as the random information, the character string associated with the target information of the specified target; the target information acquisition means acquires the target information using the authentication response voice in response to the request for the target information, The target authentication means uses voiceprint information acquired from the authentication response voice in response to the request for target information and the random information for authentication using the voiceprint information. 21. An information processing system according to the present invention. twenty three. the registration information is information that associates the voiceprint information of the subject with a question and a keyword for the question; the authentication output information includes a first question that is a question registered in the registration information, and a second question that is randomly generated in association with the first question as the random information; The authentication-related processing means authenticates the target using voiceprint information and a keyword for a first question included in the authentication response voice and the registration information, and registers the second question and the keyword for the second question as registration information of the target. 19. An information processing system according to the present invention. twenty four. The information processing device includes: further comprising a target information acquisition means for acquiring target information for identifying the target, The registration information is information that further associates the target information, The authentication-related means includes: an object specifying means for specifying the object using the acquired object information and object information included in the registration information; an object authentication means for authenticating the specified object using voiceprint information included in the authentication response voice and the registration information of the specified object, and a keyword in response to the first question; and a registration update means for registering the second question and a keyword for the second question as the registration information of the target. 23. An information processing system according to the present invention. twenty five. the authentication output information includes, together with the first question and the second question, a request for the target information as the request for the predetermined information; the target information acquisition means acquires the target information using the authentication response voice in response to the request for the target information, The target authentication means uses voiceprint information acquired from the authentication response voices to the request for target information, the first question, and the second question for authentication using the voiceprint information. 23. An information processing system according to the present invention. 26. the registration information is information that further associates frequency information of the voice in response to the first question, The authentication-related processing means performs authentication of the target by using the authentication response voice and a keyword and frequency information for the first question included in the registration information. 23. An information processing system according to the present invention. 27. One or more computers Outputting output information including randomly generated random information; acquiring a response voice in response to the output information; Processing related to authentication of the target is performed using at least one of voiceprint information and word information included in the response voice. Information processing methods. 28. Furthermore, the random information is acquired, the random information includes at least one of a second question randomly generated in relation to the first question and a randomly generated character string; The word information includes at least one of a keyword included in the response voice to the second question and the randomly generated character string. 27. Information processing method described in paragraph 2. 29. The output information further includes a predetermined request for information. 27. or 28. An information processing method. 30. In the process related to the authentication of the target, the voiceprint information and word information included in the response voice are used to perform the process related to the authentication of the target. 27. or 28. An information processing method. 31. the output information includes registration output information that is output to register the target; the response voice includes a registration response voice that is a response voice to the registration output information, The process related to authenticating the target includes registering voiceprint information and word information included in the registration response voice as registration information of the target. 29. Information processing method described in paragraph 29. 32. the output information includes authentication output information that is output to authenticate the target; the response voice includes an authentication response voice that is a response voice to the authentication output information, The process related to authenticating the target includes authenticating the target using voiceprint information and word information included in the authentication response voice and the registration information. 31. Information processing method described in paragraph 31. 33. the random information includes a randomly generated string of characters; the word information includes a character string included in the response voice in response to the random information, the registration information is information that associates the voiceprint information of the subject with the randomly generated character string; In the process related to authenticating the target, the target is authenticated using voiceprint information and character strings included in the authentication response voice and the registration information. 32. Information processing method described in paragraph 32. 34. Furthermore, target information for identifying the target is acquired, The registration information is information that further associates the target information of the target, the authentication output information includes, as the random information, the character string associated with the target information in the registration information of the target; In a process related to authentication of the object, Identifying the target using the acquired target information and target information included in the registration information; The specified target is authenticated using voiceprint information and a character string included in the authentication response voice and the registration information of the specified target. 33. Information processing method described in paragraph 33. 35. the authentication output information includes, as the request for the predetermined information, a request for the target information, and includes, as the random information, the character string associated with the target information of the specified target; In the step of acquiring the target information, the target information is acquired using the authentication response voice in response to the request for the target information; In authenticating the specified target, the voiceprint information is acquired from the request for target information and the authentication response voice in response to the random information. 34. Information processing method described in paragraph 34. 36. the registration information is information that associates the voiceprint information of the subject with a question and a keyword for the question; the authentication output information includes a first question that is a question registered in the registration information, and a second question that is randomly generated in association with the first question as the random information; In the process related to authenticating the target, the target is authenticated using voiceprint information and a keyword for a first question included in the authentication response voice and the registration information, and the second question and the keyword for the second question are registered as the registration information of the target. 32. Information processing method described in paragraph 32. 37. Furthermore, target information for identifying the target is acquired, The registration information is information that further associates the target information, In a process related to authentication of the object, Identifying the target using the acquired target information and target information included in the registration information; authenticating the identified target using voiceprint information included in the authentication response voice and the registration information of the identified target, and a keyword in response to the first question; The second question and a keyword for the second question are registered as registration information of the target. 36. Information processing method described in paragraph 36. 38. the authentication output information includes, together with the first question and the second question, a request for the target information as the request for the predetermined information; In the step of acquiring the target information, the target information is acquired using the authentication response voice in response to the request for the target information; In authenticating the specified target, the voiceprint information is acquired from the authentication response voices to the request for target information, the first question, and the second question. 36. Information processing method described in paragraph 36. 39. the registration information is information that further associates frequency information of the voice in response to the first question, In the process related to the authentication of the target, the target is authenticated using the authentication response voice and a keyword and frequency information for the first question included in the registration information. 36. Information processing method described in paragraph 36. 40. On one or more computers, A program for executing the information processing method according to any one of 27. to 39. 41. On one or more computers, A recording medium on which a program for executing the information processing method according to any one of 27 to 39 is recorded. [Explanation of symbols]

[0347] 50 devices 51 Telephone section 53 Display control unit 55 Display section 57 Input section 59 Terminal communication unit 100,200,300,400,500 Information processing equipment 110,210,410 Random information acquisition section 130,230,330,430,530 Output section 150,250,350,450,550 Audio acquisition section 170,270,370,470,570 Authentication related processing section 271,371,471,571 Registration Department 271a,371a,471a,571a Voiceprint acquisition unit 271b,471b Word Acquisition Section 271c,471c Registration control section 272,372,472,572 Authentication Section 272a Target Identification Section 272b,372b,472b,572b Target authentication section 272b1,372b1,472b1,572b1 First Authentication Section 272b2,472b2 Second Certification Section 272b3,472b3 Authentication result output section 290,390 Target information acquisition section 295,495 Registration information storage section 472c Registration and Renewal Division

Claims

1. an output means for outputting output information including randomly generated random information; a voice acquisition means for acquiring a response voice in response to the output information; and an authentication-related processing means for performing processing related to authentication of a target using at least one of voiceprint information and word information included in the response voice. Information processing device.

2. further comprising a random information acquisition means for acquiring the random information, the random information includes at least one of a second question randomly generated in relation to the first question and a randomly generated character string; The word information includes at least one of a keyword included in the response voice to the second question and the randomly generated character string. The information processing device according to claim 1 .

3. The output information further includes a predetermined request for information.

3. The information processing device according to claim 1 or 2.

4. The authentication-related processing means performs processing related to the authentication of the target using voiceprint information and word information included in the response voice.

3. The information processing device according to claim 1 or 2.

5. the output information includes registration output information that is output to register the target; the response voice includes a registration response voice that is a response voice to the registration output information, The process related to authenticating the target includes registering voiceprint information and word information included in the registration response voice as registration information of the target. The information processing device according to claim 3 .

6. the output information includes authentication output information that is output to authenticate the target; the response voice includes an authentication response voice that is a response voice to the authentication output information, The process related to authenticating the target includes authenticating the target using voiceprint information and word information included in the authentication response voice and the registration information. The information processing device according to claim 5 .

7. the random information includes a randomly generated string of characters; the word information includes a character string included in the response voice in response to the random information, the registration information is information that associates the voiceprint information of the subject with the randomly generated character string; The authentication-related processing means authenticates the target using voiceprint information and character strings included in the authentication response voice and the registration information. The information processing device according to claim 6 .

8. the registration information is information that associates the voiceprint information of the subject with a question and a keyword for the question; the authentication output information includes a first question that is a question registered in the registration information, and a second question that is randomly generated in association with the first question as the random information; The authentication-related processing means authenticates the target using voiceprint information and a keyword for a first question included in the authentication response voice and the registration information, and registers the second question and the keyword for the second question as registration information of the target. The information processing device according to claim 6 .

9. One or more computers Outputting output information including randomly generated random information; acquiring a response voice in response to the output information; Processing related to authentication of the target is performed using at least one of voiceprint information and word information included in the response voice. Information processing methods.

10. On one or more computers, Outputting output information including randomly generated random information; acquiring a response voice in response to the output information; A program for executing a process related to authentication of a target using at least one of voiceprint information and word information contained in the response voice.

Images