JP2017120576A - User authentication system, user authentication method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent voice for voiceprint authentication uttered by a user from being recorded and used fraudulently by allowing the user to input voice data easily and by permitting precise voiceprint authentication by simple means.SOLUTION: A user authentication system 1 displays a random character string generated by a character string generation part 5 on a display part 6, and inputs user voice data uttered by a user for a random character string displayed on the display part 6 from a voice data input part 2 according to a request from a voice data input request part 3. A voiceprint data generation part 8 extracts voiceprint data from user voice data for a random character string inputted from the voice input part 2, a voiceprint DB 10 stores voiceprint data generated from user voice data uttered by the user, and a voiceprint data comparison part 9 compares voiceprint data stored in the voiceprint DB 10 with voiceprint data extracted by the voiceprint data generation part 8 to output the comparison result.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a user authentication system, a user authentication method, and a program for identity verification required when a user accesses various systems.

In recent years, the spread of the Internet has been remarkable, and companies of all types of business have established websites and provide various services. Many of these websites provide services only to registered users, and in websites that are available only to registered users, whether or not they are registered users User authentication is performed to confirm the above.
This user authentication is generally performed using a user ID and a password. That is, when a user accesses the website, the user inputs a user ID for identifying the user and a password for confirming that the user is the user, and the entered user ID and password are registered. If the ID and password match, the website can be accessed (logged in).

In this case, since the password set for identity verification in the user authentication is composed of characters such as alphanumeric characters, there is a high risk of being stolen and illegally used. For this reason, as a means for user authentication, voice print authentication, which is one of biometric authentication instead of password authentication or along with password authentication, has been proposed.
For example, Japanese Patent Laying-Open No. 2005-128307 (Patent Document 1) records voice print data of elements such as numerals and characters constituting a password for each registrant, and random password voice data formed by password forming means. And a voiceprint authentication system that determines that the determination requester is a registrant by comparing voice data formed by combining the voiceprint data of each element.
In Japanese Patent Application Laid-Open No. 2007-52496 (Patent Document 2), a password and a user's voiceprint data are registered in advance, and a password registered in advance by recognizing the spoken voice of the password input by the user by voice recognition. The user who determines whether the voiceprint data of the utterance voice of the password input by the user matches the user's voiceprint data registered in advance when the password authentication is successful. An authentication system is disclosed.
Furthermore, in Japanese Patent Application Laid-Open No. 2015-99470 (Patent Document 3), a user's voice print information is registered in advance in a server, and a character string generation unit of the server generates a certain character string, which is owned by the user. To the information portable terminal, and from the information portable terminal, the voice of a certain character string read out by the user is transmitted to the server, and the voice authentication of the character string received by the authentication unit of the server from the information portable terminal is An authentication system is disclosed in which user authentication is performed based on whether or not the voiceprint information of a user registered in advance matches.

However, in the voiceprint authentication system of Patent Document 1, it is necessary to form voice data by combining voiceprint data of each element for each random password, which not only complicates the processing but also accurately prints voiceprint data of each element. There is a problem that it cannot be determined unless combined. In addition, the voiceprint authentication system of Patent Document 1 has a problem that the accuracy of the determination is deteriorated because the voiceprint combined with the voiceprints of the respective elements is collated and determined instead of the voiceprint of the entire voice uttered by the user.
Further, in the user authentication system of Patent Document 2, in order to perform voice recognition and voiceprint recognition from a user's utterance using a preset password, when the utterance of a password uttered by the user is recorded, If the recorded voice is used, the user is authenticated and there is a risk of unauthorized use.
Further, in the authentication system of Patent Document 3, since a certain character string generated on the server side is used for voiceprint recognition, as in Patent Document 2, a sound of a certain character string uttered by the user is used. There is a risk of being recorded and misused.

JP 2005-128307 A JP 2007-52496 A Japanese Patent Laying-Open No. 2015-99470

  The problem to be solved by the present invention is that when performing user authentication by voiceprint authentication, the user can easily input voiceprint data, and can perform voiceprint authentication with high accuracy by simple means. This is to prevent unauthorized use by recording the authentication voice.

  The present invention requests character string display means for displaying a character string, voice data input means for inputting user voice data issued by a user, and input of user voice data for the character string displayed on the character string display means. Voice data input request means, voice print data storage means for storing voice print data generated from user voice data issued by the user, character string generation means for generating random character strings, and the voice input request means In accordance with a request from the voice data input means, voice print data generation means for generating voice print data from user voice data for the random character string, voice print data stored in the voice print data storage means, and A user authentication system comprising voiceprint data collating means for collating voiceprint data generated by the voiceprint data generating means and outputting the result of the collation. Providing Temu solves the above problems.

  The present invention requests character string display means for displaying a character string, voice data input means for inputting user voice data issued by a user, and input of user voice data for the character string displayed on the character string display means. Voice data input requesting means, character string storage means for storing a predetermined character string, character string generating means for generating a random character string, and the voice data input means in accordance with a request from the voice input requesting means Voice print data generating means for generating voice print data from user voice data for the predetermined character string and the random character string inputted by the user, and the voice print data generating means from the user voice data for the predetermined character string Voice print data registration means for registering the generated voice print data, and the voice print data extraction means are configured to obtain user voice data for the random character string from the user voice data. To solve the above-mentioned problems by providing a user authentication system comprising voiceprint data collated with voiceprint data registered by the voiceprint data registration means and voiceprint data collation means for outputting the collation result. is there.

  The present invention recognizes a character string from user voice data for the random character string input from the voice data input means in accordance with a request from the voice input request means, and recognizes the recognized character string and the random character string. A user authentication system further comprising voice recognition means for determining whether or not a character string matches is provided to solve the above problem.

  The present invention requests character string display means for displaying a character string, voice data input means for inputting user voice data issued by a user, and input of user voice data for the character string displayed on the character string display means. A voice data input requesting means, a character string generating means for generating a random character string, and a user for the random character string input from the voice data input means in accordance with a request from the voice input requesting means. A terminal device comprising user voice data transmitting means for transmitting voice data; user voice data receiving means for receiving user voice data for the random character string transmitted by the user voice data transmitting means; and the user Voice print data storage means for storing voice print data generated from user voice data emitted by the user, and the user voice data receiving means Voice print data generation means for generating voice print data from user voice data for the random character string, voice print data stored in the voice print data storage means and voice print data generated by the voice print data generation means are collated. An object of the present invention is to solve the above problems by providing a user authentication system including a server including voiceprint data collating means for outputting the collation result.

  The present invention requests character string display means for displaying a character string, voice data input means for inputting user voice data issued by a user, and input of user voice data for the character string displayed on the character string display means. Voice data input requesting means, character string storage means for storing a predetermined character string, character string generating means for generating a random character string, and the voice data input means in accordance with a request from the voice input requesting means A terminal device comprising user voice data transmitting means for transmitting user voice data for the predetermined character string and the random character string input from the terminal, and the predetermined character transmitted by the user voice data transmitting means User voice data receiving means for receiving user voice data for a string and the random character string; and the user voice data receiving means received by the user voice data receiving means Voice print data generation means for generating voice print data from user voice data for a fixed character string and the random character string; and voice print data generated by the voice print data generation means from user voice data for the predetermined character string. The voice print data registration means to be registered, the voice print data generated by the voice print data generation means from the user voice data for the random character string, and the voice print data registered by the voice print data registration means are checked, and the result of the check Is provided to provide a user authentication system including a server including voiceprint data collating means for outputting the above.

  In the present invention, the server recognizes a character string from user voice data for the random character string received by the user voice data receiving means, and the recognized character string matches the random character string. It is an object of the present invention to provide a user authentication system further comprising voice recognition means for determining whether or not to solve the above problem.

  The present invention displays a random character string generated by the character string generation means on the character string display means, and the random character string displayed on the character string display means in accordance with a request from the voice data input request means. The user voice data issued by the user is input from the voice data input means, and the voiceprint data generation means extracts voiceprint data from the user voice data for the random character string input from the voice data input means, Voice print data generated from user voice data issued by the user in advance is stored in the voice print data storage means, and the voice print data collating means stores the voice print data stored in the voice print data storage means and the voice print data generation means. A user authentication method comprising a step of collating with extracted voiceprint data and outputting a result of the collation to solve the above problem A.

  The present invention displays the predetermined character string stored in the character string storage means on the character string display means, and the predetermined character string displayed on the character string display means in accordance with a request from the voice data input request means. User voice data issued by the user is input from voice data input means, and voice print data generation means generates voice print data from user voice data for the predetermined character string input from the voice data input means, and voice print data The registration means registers the voice print data generated from the user voice data for the predetermined character string by the voice print data generation means, displays the random character string generated by the character string generation means on the character string display means, and In response to a request from the voice data input request means, user voice data issued by the user for the random character string displayed on the character string display means. From the voice data input means, the voiceprint data generation means generates voiceprint data from user voice data for the random character string input from the voice data input means, and the voiceprint data collating means User authentication comprising the step of collating voice print data generated from user voice data for the random character string by the voice print data generating means with the voice print data registered by the voice print data registration means and outputting the result of the check A method is provided to solve the above problems.

  According to the present invention, the voice recognition means recognizes a character string from user voice data for the random character string input from the voice data input means in accordance with a request from the voice input request means, and the recognized character A user authentication method further comprising a step of determining whether a string and the random character string match is provided to solve the above problem.

  The present invention displays a random character string generated by the character string generation means on the character string display means, and the random character string displayed on the character string display means in accordance with a request from the voice data input request means. User voice data uttered by the user is input from the voice data input means, and the user voice data transmission means transmits user voice data for the random character string input from the voice data input means. A data receiving means receives user voice data for the random character string transmitted by the user voice data transmitting means, and a voiceprint data generating means for the random character string received by the user voice data receiving means. Voice print data is extracted from the user voice data, and voice print data generated from the user voice data previously issued by the user is extracted from the voice print data. The voice print data collating means collates the voice print data stored in the voice print data storage means with the voice print data generated by the voice print data generating means, and outputs the collation result. A user authentication method provided is provided to solve the above problems.

  The present invention displays the predetermined character string stored in the character string storage means on the character string display means, and the predetermined character string displayed on the character string display means in accordance with a request from the voice data input request means. User voice data issued by a user is input from voice data input means, and user voice data transmission means transmits user voice data for the predetermined character string input from the voice data input means, and receives user voice data. Means for receiving user voice data for the predetermined character string transmitted by the user voice data transmitting means, and voice print data generating means for user voice data for the predetermined character string received by the user voice data receiving means; Voice print data is generated from the voice print data registration means, and the voice print data generation means uses the user sound for the predetermined character string. The voice print data generated from the data is registered, the random character string generated by the character string generation means is displayed on the character string display means, and displayed on the character string display means in accordance with the request from the voice data input request means. Further, user voice data issued by a user for the random character string is input from the voice data input means, and the user voice data transmission means is configured to input the random character string input from the voice data input means. User voice data is transmitted, the user voice data receiving means receives user voice data for the random character string transmitted by the user voice data transmitting means, and the voiceprint data generating means is configured to receive the user voice data. Voice print data is generated from user voice data for the random character string received by the receiving means, The voiceprint data generating means collates the voiceprint data generated from the user voice data for the random character string with the voiceprint data registered by the voiceprint data registration means, and outputs the collation result. A user authentication method is provided to solve the above problems.

  In the present invention, a voice recognition unit recognizes a character string from user voice data for the random character string received by the user voice data receiving unit, and the recognized character string and the random character string are A user authentication method further comprising a step of determining whether or not they match is provided to solve the above problem.

  The present invention provides a program for causing a computer to execute the user authentication method to solve the above-described problems.

  In the user authentication system of the present invention, the random character string generated by the character string generation unit is displayed on the character string display unit, and the character string display unit displays the random character string in accordance with a request from the voice data input request unit. User voice data uttered by the user for the random character string is input from the voice data input means, and the voiceprint data generation means is used to generate a voiceprint from the user voice data for the random character string input from the voice data input means. Data is extracted, voiceprint data storage means stores voiceprint data generated from user voice data issued by the user in voiceprint data storage means, and voiceprint data collating means stores voiceprint data stored in the voiceprint data storage means And the voice print data extracted by the voice print data generation means, and the result of the check is output. Upon performing testimony, users easily can enter voiceprint data, an effect that it is accurate voiceprint authentication by simple means.

  In the user authentication system of the present invention, the predetermined character string stored in the character string storage unit is displayed on the character string display unit, and the character string display unit displays the predetermined character string in accordance with a request from the voice data input request unit. User voice data uttered by the user for a predetermined character string is input from the voice data input means, and voiceprint data generation means obtains voiceprint data from the user voice data for the predetermined character string input from the voice data input means. The voice print data registration means generates the voice print data generated from the user voice data for the predetermined character string, and the random character string generated by the character string generation means is the character string display means. The random character string displayed on the character string display means according to the request from the voice data input request means User voice data issued by the user is input from voice data input means, and the voice print data generation means generates voice print data from user voice data for the random character string input from the voice data input means. The voiceprint data collating means collates the voiceprint data generated from the user voice data for the random character string by the voiceprint data generating means with the voiceprint data registered by the voiceprint data registering means, Therefore, when performing user authentication by voiceprint authentication, the user can easily register and input voiceprint data, and it is possible to perform highly accurate voiceprint authentication with simple means.

  In the user authentication system of the present invention, in accordance with a request from the voice input request unit, a character string is recognized from user voice data for the random character string input from the voice data input unit, and the recognized character is Since the apparatus further comprises voice recognition means for determining whether the sequence and the random character string match, it is possible to further prevent the voiceprint authentication voice issued by the user from being recorded and illegally used. Play.

  In the user authentication system of the present invention, on the terminal device side, a random character string generated by the character string generation means is displayed on the character string display means, and the character string display means is in accordance with a request from the voice data input request means User voice data issued by the user for the random character string displayed on the screen is input from the voice data input means, and the user voice data transmission means is configured to input the random character string input from the voice data input means. The user voice data receiving means receives the user voice data for the random character string sent by the user voice data sending means, and the voiceprint data generating means Voice print data is extracted from user voice data for the random character string received by the voice data receiving means, and the user Voice print data generated from the generated user voice data is stored in the voice print data storage means, and the voice print data collating means stores the voice print data stored in the voice print data storage means and the voice print data generated by the voice print data generation means. And when the user authentication is performed by voiceprint authentication using the communication line, the user can easily input the voiceprint data, and the voiceprint authentication can be performed with a simple means with high accuracy. There is an effect.

  In the user authentication system of the present invention, on the terminal device side, a predetermined character string stored in the character string storage means is displayed on the character string display means, and the character string display means is in accordance with a request from the voice data input request means. The user voice data uttered by the user for the predetermined character string displayed on the screen is input from the voice data input means, and the user voice data transmitting means is the user for the predetermined character string input from the voice data input means. Voice data is transmitted, and on the server side, user voice data reception means receives user voice data for the predetermined character string transmitted by the user voice data transmission means, and voice print data generation means receives the user voice data. Voice print data is generated from user voice data for the predetermined character string received by the means, and voice print data registration means Voiceprint data generating means registers voiceprint data generated from user voice data for the predetermined character string, and the terminal device side uses the random character string generated by the character string generating means as character string display means. In response to a request from the voice data input request means, user voice data issued by a user for the random character string displayed on the character string display means is input from the voice data input means, and the user voice Data transmission means transmits user voice data for the random character string input from the voice data input means, and on the server side, the user voice data reception means is transmitted by the user voice data transmission means. User voice data for the random character string is received, and the voiceprint data generating means receives the user voice data Voiceprint data is generated from user voice data for the random character string received by the stage, and voiceprint data collating means generates voiceprint data generated by the voiceprint data generating means from the user voice data for the random character string. And the voiceprint data registered by the voiceprint data registration means, and the collation result is output, so that the user can easily register and input the voiceprint data when performing user authentication by voiceprint authentication using a communication line. Thus, it is possible to perform highly accurate voiceprint authentication with simple means.

  In the user authentication system of the present invention, the server recognizes a character string from user voice data for the random character string received by the user voice data receiving means, and recognizes the recognized character string and the random character string. Since voice recognition means for determining whether or not the character strings match is further provided, the voice print authentication voice produced by the user can be prevented from being recorded and illegally used.

  In the user authentication method of the present invention, the random character string generated by the character string generation unit is displayed on the character string display unit, and the character string display unit displays the random character string in accordance with a request from the voice data input request unit. User voice data uttered by the user for the random character string is input from the voice data input means, and the voiceprint data generation means is used to generate a voiceprint from the user voice data for the random character string input from the voice data input means. Data is extracted, voice print data generated from user voice data issued by the user in advance is stored in the voice print data storage means, and voice print data collating means and the voice print data stored in the voice print data storage means The voiceprint data generation means includes a step of collating with the voiceprint data extracted and outputting the collation result. When performing user authentication, the user can easily enter the voiceprint data, an effect that it is accurate voiceprint authentication by simple means.

  In the user authentication method of the present invention, the predetermined character string stored in the character string storage means is displayed on the character string display means, and the character string display means displays the predetermined character string in accordance with a request from the voice data input request means. User voice data uttered by the user for a predetermined character string is input from the voice data input means, and voiceprint data generation means obtains voiceprint data from the user voice data for the predetermined character string input from the voice data input means. The voice print data registration means generates the voice print data generated from the user voice data for the predetermined character string, and the random character string generated by the character string generation means is the character string display means. The random character string displayed on the character string display means according to a request from the voice data input request means Voice data input means is input from voice data input means, and the voiceprint data generation means generates voiceprint data from user voice data for the random character string input from the voice data input means, The data collating unit collates the voice print data generated from the user voice data for the random character string by the voice print data generating unit and the voice print data registered by the voice print data registering unit, and outputs the collation result. Since the steps are provided, when performing user authentication by voiceprint authentication, the user can easily register and input voiceprint data, and high-precision voiceprint authentication can be performed with simple means.

  In the user authentication method of the present invention, the voice recognition unit recognizes a character string from user voice data for the random character string input from the voice data input unit in accordance with a request from the voice input request unit. Since the method further comprises the step of determining whether the recognized character string matches the random character string, it is further possible to prevent the voiceprint authentication voice uttered by the user from being recorded and illegally used. There is an effect.

  In the user authentication method of the present invention, the random character string generated by the character string generation unit is displayed on the character string display unit, and the character string display unit displays the random character string in accordance with a request from the voice data input request unit. User voice data issued by the user for the random character string is input from the voice data input means, and the user voice data transmission means receives the user voice data for the random character string input from the voice data input means. And the user voice data receiving means receives the user voice data for the random character string sent by the user voice data sending means, and the voice print data generating means is received by the user voice data receiving means. Voice print data is extracted from user voice data for an artificial character string, and generated from user voice data previously issued by the user The voiceprint data is stored in the voiceprint data storage means, and the voiceprint data collating means collates the voiceprint data stored in the voiceprint data storage means with the voiceprint data generated by the voiceprint data generating means, Since the method includes a step of outputting a result, when performing user authentication by voiceprint authentication using a communication line, the user can easily input voiceprint data, and high-precision voiceprint authentication can be performed with simple means. Play.

  In the user authentication method of the present invention, the predetermined character string stored in the character string storage means is displayed on the character string display means, and the character string display means displays the predetermined character string in accordance with a request from the voice data input request means. User voice data uttered by a user for a predetermined character string is input from voice data input means, and user voice data transmission means transmits user voice data for the predetermined character string input from the voice data input means. The user voice data receiving means receives the user voice data for the predetermined character string transmitted by the user voice data transmitting means, and the voice print data generating means is the predetermined character string received by the user voice data receiving means. Voice print data is generated from user voice data for the voice print data registration means, and the voice print data generation means The voice print data generated from the user voice data for the character string is registered, the random character string generated by the character string generation means is displayed on the character string display means, and the character is displayed according to the request from the voice data input request means. User voice data issued by the user for the random character string displayed on the column display means is input from the voice data input means, and the user voice data transmission means is input from the voice data input means. User voice data for a random character string is transmitted, and the user voice data receiving means receives user voice data for the random character string transmitted by the user voice data transmitting means, and the voiceprint data generating means However, voiceprint data is obtained from user voice data for the random character string received by the user voice data receiving means. The voiceprint data collating means collates the voiceprint data generated from the user voice data for the random character string by the voiceprint data generating means with the voiceprint data registered by the voiceprint data registering means, and Since the step of outputting the result is provided, when performing user authentication by voiceprint authentication using a communication line, the user can easily register and input voiceprint data, and high-precision voiceprint authentication can be performed by a simple means. There is an effect.

  In the user authentication method of the present invention, the voice recognition means recognizes a character string from user voice data for the random character string received by the user voice data receiving means, and the recognized character string and Since the method further includes the step of determining whether or not the intentional character string matches, there is an effect that it is possible to prevent the voiceprint authentication voice uttered by the user from being recorded and illegally used.

  In the program of this invention, there exists the same effect as the effect which the said user authentication method show | plays.

It is a block diagram which shows the structure of 1st Embodiment of the user authentication system of this invention. It is a block diagram which shows the structure of 2nd Embodiment of the user authentication system of this invention. It is a block diagram which shows the structure of 3rd Embodiment of the user authentication system of this invention. It is a block diagram which shows the structure of 4th Embodiment of the user authentication system of this invention. 3 is a flowchart showing the operation of the user authentication system 1. FIG. 10 is an explanatory diagram illustrating an example of a display screen of the display unit 6 when the voice data input request unit 3 requests a voice input from the user in the authentication process by the user authentication system 1. FIG. 11 is an explanatory diagram illustrating an example of a display screen of the display unit 6 when the voice data input unit 2 is inputting voice data in the authentication process by the user authentication system 1. 7 is an explanatory diagram illustrating an example of a display screen of the display unit 6 from the voice recognition by the voice recognition unit 7 to the output of the collation result by the voiceprint data collation unit 9 in the authentication process by the user authentication system 1; FIG. FIG. 10 is an explanatory diagram illustrating an example of a screen 110 of the display unit 6 after login processing is performed by the login unit 101 of the external system 100. 5 is a flowchart showing voiceprint data registration processing by the user authentication system 20. FIG. 10 is an explanatory diagram showing an example of a display screen of the display unit 26 when a voice data input request unit 23 requests voice input from a user in a voiceprint data registration process by the user authentication system 20. FIG. 11 is an explanatory diagram showing an example of a display screen of the display unit 26 when a voice data input unit 22 is inputting voice data in voiceprint data registration processing by the user authentication system 20. 7 is an explanatory view showing an example of a display screen of the display unit 26 from the generation of voice print data by the voice print data generation unit 28 to the registration of voice print data by the voice print data registration unit 31 in the voice print data registration process by the user authentication system 20. FIG. . FIG. 10 is an explanatory diagram showing an example of a display screen of the display unit 26 when registration of voiceprint data by the voiceprint data registration unit 31 is completed in the voiceprint data registration process by the user authentication system 20. 5 is a flowchart showing the operation of the user authentication system 40. 5 is a flowchart showing voiceprint data registration processing by a user authentication system 70;

[User authentication system configuration]
First, the configuration of the user authentication system of the present invention will be described.
FIG. 1 is a block diagram showing the configuration of the first embodiment of the user authentication system of the present invention, FIG. 2 is a block diagram showing the configuration of the second embodiment of the user authentication system of the present invention, and FIG. FIG. 4 is a block diagram showing the configuration of the third embodiment of the user authentication system of the present invention, and FIG. 4 is a block diagram showing the configuration of the fourth embodiment of the user authentication system of the present invention.
In FIG. 1, 1 is a user authentication system, 2 is a voice data input unit, 3 is a voice data input request unit, 4 is a voice data storage unit, 5 is a character string generation unit, and 6 is a character string display means of the present invention. A display unit, 7 is a voice recognition unit, 8 is a voice print data generation unit, 9 is a voice print data collation unit, and 10 is a voice print database (DB) serving as voice print data storage means of the present invention.
In FIG. 2, 20 is a user authentication system, 21 is a character string storage unit, 22 is a voice data input unit, 23 is a voice data input request unit, 24 is a voice data storage unit, 25 is a character string generation unit, and 26 is the present invention. 27 is a voice recognition unit, 28 is a voiceprint data generation unit, 29 is a voiceprint data collation unit, 30 is a voiceprint database (DB) as voiceprint data storage means of the present invention, and 31 is a voiceprint. A data registration unit.
In FIG. 3, 40 is a user authentication system, 41 is a terminal device, 42 is a voice data input unit, 43 is a voice data input request unit, 44 is a voice data storage unit, 45 is a character string generation unit, and 46 is a character of the present invention. Display unit that is a column display means, 47 is a transmission unit, 48 is a reception unit, 50 is a server, 51 is a reception unit, 52 is a transmission unit, 54 is a data storage unit, 57 is a voice recognition unit, and 58 is a voiceprint data generation unit , 59 is a voiceprint data collation unit, 60 is a voiceprint database (DB) serving as voiceprint data storage means of the present invention, and 65 and 66 are communication lines.
In FIG. 4, 70 is a user authentication system, 71 is a terminal device, 72 is a voice data input unit, 73 is a voice data input request unit, 74 is a voice data storage unit, 75 is a character string generation unit, and 76 is a character of the present invention. Display unit as column display means, 77 is a character string storage unit, 78 is a transmission unit, 79 is a reception unit, 80 is a server, 81 is a reception unit, 82 is a transmission unit, 84 is a data storage unit, and 87 is a voice recognition unit , 88 is a voiceprint data generation unit, 89 is a voiceprint data collation unit, 90 is a voiceprint database (DB) serving as voiceprint data storage means of the present invention, and 91 is a voiceprint data registration unit.
1-4, 100 is an external system, 101 is the login part.

[User authentication system 1]
As shown in FIG. 1, a user authentication system 1 according to a first embodiment of the present invention includes a voice data input unit 2, a voice data input request unit 3, a voice data storage unit 4, a character string generation unit 5, and a display unit. 6, voice recognition unit 7, voice print data generation unit 8, voice print data collation unit 9, voice print DB 10, and the like.

The audio data input unit 2 is composed of a microphone and its control device and the like, and takes in the audio uttered by the user as audio data. In this case, the voice data input unit 2 captures a voice uttered by the user as voice data for the random character string generated by the character string generation unit 5.
The voice data input request unit 3 causes the display unit 6 to display characters or the like for requesting voice input from the user so that the user utters voice and captures it from the voice input unit 2.
The voice data storage unit 4 temporarily stores the voice data (voice data for a random character string) input from the voice data input unit 2, and then stores the saved voice data into the voice recognition unit 7 and the voiceprint data generation unit 8. Send to.
The character string generation unit 5 generates a random character string, for example, a random number, and sends the generated random character string to the display unit 6 and the voice recognition unit 7.
The display unit 6 is composed of a display, its control device, and the like. The display unit 6 receives characters from the voice data input request unit 3 for requesting voice input, random character strings and voiceprint data generated by the character string generation unit 5. The collation result sent from the collation unit 9 is displayed.

The voice recognition unit 7 includes a character string recognition unit and a character string determination unit (not shown).
The character string recognizing unit performs acoustic analysis on the voice data (sound data on a random character string) sent from the voice data storage unit 4, converts the voice data into kana characters, and pronounces the kana characters. Using a dictionary or the like, it is converted into a character string in Japanese, and what kind of character string the speech data consists of is recognized.
The character string determining unit determines whether the character string recognized by the character string recognizing unit matches the random character string sent from the character string generating unit 5, and the determination result (successful speech recognition when matching) , The voice recognition data unsuccessful is sent to the voiceprint data generation unit 8.

When the voice recognition data generation unit 8 receives a determination result indicating that the voice recognition is successful from the voice recognition unit 7, the voice print data generation unit 8 receives the voice data (voice data for a random character string) sent from the voice data storage unit 4. ) To generate voiceprint data represented by a spectrogram (a three-dimensional graph showing the strength of the time, frequency, and signal component) that indicates how much sound and what frequency is included at what time. The generated voice print data is sent to the voice print data collation unit 9.
The voiceprint data collation unit 9 collates the voiceprint data sent from the voiceprint data generation unit 8 with the user's voiceprint data stored in the voiceprint DB 10, and the difference is small and within a certain range (below the threshold). For example, a collation result indicating that the voiceprint data is collated successfully (authentication success) is output.
The voiceprint DB 10 stores voiceprint data (voiceprint data specified by a user ID) generated in advance from voice data of a user who uses the user authentication system 1.

[User authentication system 20]
As shown in FIG. 2, the user authentication system 20 according to the second embodiment of the present invention includes a character string storage unit 21, a voice data input unit 22, a voice data input request unit 23, a voice data storage unit 24, a character string. The generation unit 25, the display unit 26, the voice recognition unit 27, the voice print data generation unit 28, the voice print data collation unit 29, the voice print DB 30, the voice print data registration unit 31, and the like.
This user authentication system 20 is different from the configuration of the user authentication system 1 in that it includes a character string storage unit 21 and a voiceprint data registration unit 31 in addition to the configuration of the user authentication system 1.

The character string storage unit 21 stores a predetermined character string, for example, a 10-digit number in which 0 to 9 are consecutive.
The voice data input unit 22 is configured by a microphone and its control device and the like, similar to the voice data input unit 2, and takes in a voice uttered by the user as voice data. In this case, the voice data input unit 22 takes in a voice uttered by the user as voice data for a predetermined character string stored in the character string storage unit 21 and a random character string generated by the character string generation unit 5.
The voice data input request unit 23 causes the display unit 26 to display characters or the like for requesting voice input from the user in order for the user to utter a voice and capture it from the voice input unit 22.
The voice data storage unit 24 temporarily stores the voice data input from the voice data input unit 22 (voice data for a predetermined character string or a random character string), and then the stored voice data is a random character string. If so, the voice data for the random character string is sent to the voice recognition unit 27 and the voiceprint data generation unit 28. If the stored voice data is a predetermined character string, the voice data for the predetermined character string is transmitted. Is sent to the voiceprint data generation unit 28.
The character string generation unit 25 generates a random character string, for example, a random number, and sends the generated random character string to the display unit 26 and the voice recognition unit 27.
The display unit 26 includes a display and its control device. The display unit 26 receives characters from the voice data input request unit 23 for requesting voice input, a predetermined character string stored in the character string storage unit 21, and a character string. A random character string generated by the generation unit 25, a collation result sent from the voiceprint data collation 29, and the like are displayed.

Similar to the voice recognition unit 7, the voice recognition unit 27 includes a character string recognition unit and a character string determination unit (not shown).
The character string recognizing unit has the same function as the character string recognizing unit of the voice recognizing unit 7, and the voice data of the random character string sent from the voice data storing unit 24 is composed of any character string. Recognize.
The character string determination unit has the same function as the character string determination unit of the speech recognition unit 7, and the character string recognized by the character string recognition unit matches the random character string sent from the character string generation unit 25. The determination result is sent to the voiceprint data generation unit 28. The result of the determination (successful voice recognition succeeds, and the unmatched voice recognition fails).

The voiceprint data generation unit 28 generates voiceprint data from the voice data (speech data about a predetermined character string or random character string) sent from the voice data storage unit 24 in the same manner as the voiceprint data generation unit 8. However, the processing differs depending on the type of audio data.
That is, when voice data for a predetermined character string is sent from the voice data storage unit 24, the voiceprint data generation unit 28 generates voiceprint data from the voice data, and uses the generated voiceprint data as the voiceprint data registration unit 31. Send to.
Further, when voice data for a random predetermined character string is sent from the voice data storage unit 24 and a determination result that voice recognition is successful is sent from the voice recognition unit 27, a voiceprint The data generation unit 28 generates voiceprint data from the voice data, and sends the generated voiceprint data to the voiceprint data matching unit 29.
The voiceprint data registration unit 31 uses the voiceprint data (voiceprint data generated from the voice data for a predetermined character string) sent from the voiceprint data generation unit 28 as the voiceprint data (user ID) of the user who issued the voice. It is registered in the voiceprint DB 30 as specified voiceprint data).
The voiceprint data collation unit 29 collates the voiceprint data (voiceprint data generated from the voice data for a random character string) sent from the voiceprint data generation unit 28 with the user's voiceprint data stored in the voiceprint DB 30. If the difference is small and within a certain range (below the threshold value), a collation result indicating that the voiceprint data has been successfully collated (authentication success) is output.
The voiceprint DB 30 stores voiceprint data (voiceprint data specified by a user ID) registered by the voiceprint data registration unit 31.

[User authentication system 40]
As shown in FIG. 3, the user authentication system 40 according to the third embodiment of the present invention includes a terminal device 41, a server 50, and communication lines 65, 66 connecting them.

[Terminal device 41]
The terminal device 41 includes an audio data input unit 42, an audio data input request unit 43, an audio data storage unit 44, a character string generation unit 45, a display unit 46, a transmission unit 47, a reception unit 48, and the like.
The functions of the voice data input unit 42 and the voice data input request unit 43 are the same as the functions of the voice data input unit 2 and the voice data input request unit 3 of the user authentication system 1.

The voice data storage unit 44 temporarily stores the voice data input from the voice data input unit 42 (voice data for a random character string), and then sends the saved voice data to the transmission unit 47.
The character string generation unit 45 generates a random character string, for example, a random number, and sends the generated random character string to the display unit 46 and the transmission unit 47.
The display unit 46 is composed of a display and its control device and the like. The display unit 46 receives characters from the voice data input request unit 43 for requesting voice input, a random character string generated by the character string generation unit 45, and a reception unit. 48 displays the collation result of the voiceprint data collation unit 59 received.
The transmission unit 47 sends the voice data from the voice data storage unit 44 (voice data for a random character string) and the random character string from the character string generation unit 45 to the server 50 via the communication line 65. .

[Server 50]
The server 50 includes a reception unit 51, a transmission unit 52, a data storage unit 54, a voice recognition unit 57, a voice print data generation unit 58, a voice print data collation unit 59, a voice print DB 60, and the like.

The receiving unit 51 receives voice data (voice data for a random character string) and a random character string transmitted from the transmission unit 47 of the terminal device 41, and sends the received data to the data storage unit 54.
The data storage unit 54 temporarily stores the voice data (speech data about a random character string) and a random character string sent from the receiving unit 51, and then stores the stored voice data and the random character string. The data is sent to the voice recognition unit 57, and only the voice data is sent to the voice print data generation unit 58.

Similar to the speech recognition unit 7, the speech recognition unit 57 includes a character string recognition unit and a character string determination unit (not shown), and the character string recognition unit is the same as the character string recognition unit of the speech recognition unit 7. It has a function and recognizes what kind of character string the voice data of the random character string sent from the data storage unit 54 consists of.
The character string determination unit has the same function as the character string determination unit of the speech recognition unit 7, and the character string recognized by the character string recognition unit and the random character string sent from the data storage unit 54 It is determined whether or not they match, and the determination result (successful speech recognition is determined when they match, and unsuccessful speech recognition when they do not match) is sent to the voiceprint data generation unit 58.

The voiceprint data generation unit 58 has the same function as the voiceprint data generation unit 8. When the determination result that the voice recognition is successful is sent from the voice recognition unit 57, it is sent from the data storage unit 54. Voice print data is generated from the voice data (voice data for a random character string), and the generated voice print data is sent to the voice print data matching unit 59.
The voiceprint data collation unit 59 collates the voiceprint data sent from the voiceprint data generation unit 58 with the user's voiceprint data stored in the voiceprint DB 60, and the difference is small and within a certain range (below the threshold). For example, a collation result indicating that the voiceprint data is collated successfully (authentication success) is output.
The transmission unit 52 transmits the collation result sent from the voiceprint data collation unit 59 to the terminal device 41 via the communication line 66.
The voiceprint DB 60 stores voiceprint data (voiceprint data specified by a user ID) generated in advance from voice data of a user who uses the user authentication system 40.

[User authentication system 70]
As shown in FIG. 4, a user authentication system 70 according to the fourth embodiment of the present invention includes a terminal device 71, a server 80, and communication lines 95 and 96 connecting these.

[Terminal device 71]
The terminal device 71 includes an audio data input unit 72, an audio data input request unit 73, an audio data storage unit 74, a character string generation unit 75, a display unit 76, a character string storage unit 77, a transmission unit 78, a reception unit 79, and the like. Is done.
This terminal device 71 is different from the configuration of the terminal device 41 in that it includes a character string storage unit 77 in addition to the configuration of the terminal device 41 of the user authentication system 40.

The functions of the voice data input unit 72 and the voice data input request unit 73 are the same as the functions of the voice data input unit 22 and the voice data input request unit 23 of the user authentication system 20.
That is, the voice data input unit 72 takes in a voice uttered by the user as voice data for a predetermined character string stored in the character string storage unit 77 and a random character string generated by the character string generation unit 75.
The voice data storage unit 74 temporarily stores the voice data input from the voice data input unit 42 (voice data about a predetermined character string or a random character string), and then sends the stored voice data to the transmission unit 47. .
The character string generation unit 75 generates a random character string, for example, a random number, and sends the generated random character string to the display unit 76 and the transmission unit 78.
The display unit 76 includes a display and its control device. The display unit 76 receives characters from the voice data input requesting unit 73 that require voice input, random characters generated by the character string generating unit 75, and character strings. A predetermined character string stored in the storage unit 77, a collation result of the voiceprint data collation unit 89 received by the reception unit 79, and the like are displayed.
The transmission unit 78 sends the voice data from the voice data storage unit 74 (voice data for a predetermined character string or a random character string) and the random character string from the character string generation unit 75 via the communication line 95. To the server 80.

[Server 80]
The server 80 includes a reception unit 81, a transmission unit 82, a data storage unit 84, a voice recognition unit 87, a voice print data generation unit 88, a voice print data collation unit 89, a voice print DB 90, a voice print data registration unit 91, and the like.
This server 80 is different from the configuration of the server 50 in that it includes a voiceprint data registration unit 91 in addition to the configuration of the server 50 of the user authentication system 40.

The receiving unit 81 receives voice data (speech data for a predetermined character string or random character string) and a random character string transmitted from the transmission unit 78 of the terminal device 71, and stores the received data as data. Send to part 84.
The data storage unit 84 temporarily stores voice data (speech data for a predetermined character string or random character string) and a random character string sent from the receiving unit 81, and then stores the stored voice data. If the voice data is about a random character string, the voice data is sent to the voice recognition unit 87 and the voiceprint data generation unit 88, and the random character string is sent to the voice recognition unit 87. If the voice data sent from the receiver 81 is voice data for a predetermined character string, the voice data is sent to the voiceprint data generator 88.

Similar to the speech recognition unit 7, the speech recognition unit 87 includes a character string recognition unit and a character string determination unit (not shown), and the character string recognition unit is the same as the character string recognition unit of the speech recognition unit 7. It has a function and recognizes what kind of character string the voice data of the random character string sent from the data storage unit 84 is composed of.
The character string determination unit has the same function as the character string determination unit of the speech recognition unit 7, and the character string recognized by the character string recognition unit and the random character string sent from the data storage unit 84 are It is determined whether or not they match, and the determination result (successful speech recognition is determined if they match, and unsuccessful speech recognition is determined if they do not match) is sent to the voiceprint data generation unit 88.

The voiceprint data generation unit 88 has the same function as the voiceprint data generation unit 28. When voice data for a predetermined character string is sent from the data storage unit 84, the voiceprint data generation unit 88 uses the voice data. Voice print data is generated, and the generated voice print data is sent to the voice print data registration unit 91.
In addition, when voice data for a random predetermined character string is sent from the data storage unit 84 and a determination result that the voice recognition is successful is sent from the voice recognition unit 87, voiceprint data The generation unit 88 generates voiceprint data from the voice data, and sends the generated voiceprint data to the voiceprint data matching unit 89.
The voiceprint data registration unit 91 has the same function as the voiceprint data registration unit 31, and the voiceprint data sent from the voiceprint data generation unit 88 (voiceprint data generated from voice data for a predetermined character string) It is registered in the voiceprint DB 90 as voiceprint data (voiceprint data specified by the user ID) of the user who uttered the voice.
The voiceprint data collation unit 89 collates the voiceprint data sent from the voiceprint data generation unit 88 (voiceprint data generated from voice data for a random character string) and the user's voiceprint data stored in the voiceprint DB 90. If the difference is small and within a certain range (below the threshold value), a collation result indicating that the voiceprint data has been successfully collated (authentication success) is output.
The transmitter 82 transmits the collation result sent from the voiceprint data collator 89 to the terminal device 71 via the communication line 96.
The voiceprint DB 90 stores voiceprint data (voiceprint data specified by the user ID) registered by the voiceprint data registration unit 91.

[External system]
An external system connected to the user authentication systems 1, 20, 40, and 70 will be described.
The external system 100 is a website provided on the Internet by a company, for example, an account opening website provided by a bank, a securities company, an insurance company, etc., and provides a user with a specific service such as opening an account. provide.
The external system 100 is connected to the servers 50 and 80 in the user authentication systems 40 and 70.
Then, the login unit 101 of the external system 100 receives a collation success (authentication success) collation result from the voice print data collation units 9, 29, 59, 89 of the user authentication systems 1, 20, 40, 70. A login process for permitting the user to use the external system 100 is performed.
In addition, the display units 6, 26, 46, and 79 of the user authentication systems 1, 20, 40, and 70 are display units of an external system.

[Operation of user authentication system]
Next, operations of the user authentication systems 1, 20, 40, and 70 will be described with specific examples.
In this case, the user authentication system 1 and the user authentication system 40 perform only an authentication process, and the user authentication system 20 and the user authentication system 40 perform a voiceprint data registration process and an authentication process.

[Operation of User Authentication System 1]
FIG. 5 is a flowchart showing the operation of the user authentication system 1, and FIGS. 6 to 8 are explanatory diagrams showing examples of screens displayed on the display unit 6 in the authentication process by the user authentication system 1.
The operation of the user authentication system 1, that is, the authentication process in the user authentication system 1 will be specifically described below with reference to the drawings.

When the user authentication system 1 is activated and a user ID is input, first, the character string generation unit 5 generates a random character string (S1).
Specifically, random alphanumeric characters, symbols, kana characters, and the like are generated. Here, a 5-digit random number is generated. The 5-digit random number (number) generated by the character string generation unit 5 is different every time.
Next, the random character string generated by the character string generation unit 5 is displayed on the display unit 6, and the voice data input request unit 3 requests the user to input voice (S2).
FIG. 6 shows a display screen of the display unit 6 when the voice data input request unit 3 requests the user to input voice. In the figure, 110 is a display screen, 111 is a user ID display section, 112 is a random number display section, 113 is a start button, and 114 is a sentence display section.
As shown in FIG. 6, on the display screen 110, the user ID “00001” input by the user is displayed on the upper user ID display unit 111, and the character string generation unit 5 generates the random number display unit 112 below it. A five-digit random number is displayed together with the reading, a start button 113 is displayed below it, and a text “start button” is displayed on the text display portion 114 below that indicates that the voice data input request unit 3 requests the user to input voice. Please tell me the number. "Is displayed.

Next, the user presses the start button 113 according to the text displayed on the text display unit 114, and the numbers “1”, “3”, “5”, “6”, “4” displayed on the random number display unit 112 are displayed. When the readings “1”, “san”, “go”, “roku”, “yon” are uttered by voice, the voice data input unit 2 inputs voice data issued by the user (S3).
FIG. 7 shows a display screen of the display unit 6 when the audio data input unit 2 is inputting audio data. In the figure, 115 is a text display portion, 116 is an end button, and 117 is a cancel button.
As shown in FIG. 7, in the text display section 115 in the middle of the display screen, the voice data input request section 3 requests the user to continue voice input, etc. “Please speak slowly. Finish after reading the number. "Please press" is displayed, and an end button 116 and a cancel button 117 are displayed below it.
When the user finishes pronunciation of the number displayed on the random number display unit 112, the user presses the end button 116, thereby completing the voice input from the voice data input unit 2. Note that the user presses the cancel button 117 when canceling the voice input.
The voice data storage unit 4 temporarily stores the user voice data input by the voice data input unit 2 as described above (S4).

Next, the character string recognizing unit of the voice recognizing unit 7 recognizes what character string the voice data (voice data of five digits) sent from the voice data storage unit 4 is composed of (S5). .
Next, the character string determination unit of the voice recognition unit 7 performs a character string recognized by the character string recognition unit and a random character string (5-digit random number (number)) generated by the character string generation unit 5 (this number is a character It is determined whether the data is sent from the column generator 5 to the speech recognizer 7 (S6).
If it is determined in step S6 that they match, the determination result (successful voice recognition) is sent to the voiceprint data generation unit 8, and the process proceeds to the next voiceprint authentication step (S7). The process proceeds to step S9 without performing voiceprint authentication.

If it is determined in step S6 that they match, the voice print data generation unit 8 generates voice print data from the voice data (voice data for a 5-digit random number (number)) sent from the voice data storage unit 4. (S7).
Next, the voiceprint data matching unit 9 sends the voiceprint data generated by the voiceprint data generation unit 8 (this voiceprint data is sent from the voiceprint data generation unit 8 to the voiceprint data matching unit 9) and the user's voice stored in the voiceprint DB 10. The voiceprint data (voiceprint data specified by the user ID “00001”) is collated (S8).
If the difference between the voiceprint data to be compared in the step S8 is small and within a certain range (below the threshold value), a verification result indicating that the voiceprint data has been successfully verified (authentication success) is output. If the difference between the voiceprint data to be collated is large and outside a certain range (greater than or equal to the threshold value), a collation result indicating that the voiceprint data is unsuccessfully collated (authentication unsuccessful) is output (S9), and the process ends.
If it is determined in step S6 that the voice data does not match, voiceprint authentication (S7, S8) is not performed, a verification result of unsuccessful authentication is output (S9), and the process ends.
FIG. 8 shows a display screen of the display unit 6 from the voice recognition by the voice recognition unit 7 (S5) to the output of the collation result by the voiceprint data collation unit 9 (S9). In the figure, 118 is a character display section.
As shown in FIG. 8, from the step (S5) in which the speech recognition unit 7 recognizes what character string the speech data (speech data for a 5-digit random number (number)) consists of, the voiceprint data matching unit 9 Until the step of outputting the collation result (S9), the character “authenticating” is displayed on the character display portion 118 in the middle of the screen 110.

  As described above, since the random character string (5-digit random number) generated by the character string generation unit 5 is different every time, the user's voice data input by the voice data input unit 2 is also different every time. Even if is recorded and used for authentication, the speech recognition is unsuccessful in step S6, and a verification result of unsuccessful authentication is output, thereby preventing unauthorized use.

[Connection with external system 100]
In the collation result output from the voiceprint data collation unit 9 in step S9, if the collation result is a successful authentication, the collation result is sent to the login unit 101 of the external system 100 and a login process is performed.
FIG. 9 is an explanatory diagram showing an example of the screen 110 of the display unit 6 after the login process is performed by the login unit 101 of the external system 100, in which 119 is a text display unit.
As shown in FIG. 9, a text such as “Welcome to AA Company” is displayed on the text display unit 119 on the display screen, and the user can use various services (not shown) through the external site 100. .

[Operation of User Authentication System 20]
As described above, the user authentication system 20 performs voiceprint data registration processing and authentication processing, and each of them will be described.

[Voiceprint data registration process by user authentication system 20]
FIG. 10 is a flowchart showing a voice print data registration process by the user authentication system 20. FIGS. 11 to 14 show examples of screens displayed on the display unit 6 in the voice print data registration process by the user authentication system 20. It is explanatory drawing which showed.
The voice print data registration process by the user authentication system 20 will be specifically described below with reference to the drawings.

When the user authentication system 20 is activated and a user ID is input, first, a predetermined character string stored in the character string storage unit 21 is displayed on the display unit 26, and the voice data input request unit 23 inputs voice to the user. Is requested (S10).
The character string storage unit 21 stores predetermined character strings such as alphanumeric characters, symbols, and kana characters, but here, consecutive numbers from 0 to 9 are stored. These consecutive numbers from 0 to 9 are displayed.
FIG. 11 shows a display screen of the display unit 26 when the voice data input request unit 23 requests the user to input voice. In the figure, 120 is a display screen, 121 is a character display section, 122 is a number display section, 123 is a registration start button, and 124 is a sentence display section.
As shown in FIG. 11, on the display screen 120, the character “Voice Registration” indicating that voiceprint registration is performed is displayed on the upper character display unit 121, and the character string generation unit 5 is displayed on the number display unit 122 below it. The stored numbers from 0 to 9 are displayed together with the reading, a registration start button 123 is displayed below the number, and the voice data input request unit 23 inputs voice to the user in the text display unit 124 below the button. The text “Request a registration start button and speak slowly.” Is displayed.

Next, the user presses the registration start button 123 in accordance with the text displayed on the text display unit 124, and consecutive numbers “0”, “1”, “2”, 0 to 9 displayed on the number display unit 122, When reading “3”,..., “9” “Zero” “Ichi”, “Ni”,..., “Kyu” by voice, the voice data input unit 22 is issued by the user. Audio data is input (S11).
FIG. 12 shows a display screen of the display unit 26 when the audio data input unit 22 is inputting audio data. In the figure, 125 is a text display section, and 126 is an end button.
As shown in FIG. 12, in the text display section 125 in the middle of the display screen 120, the voice data input requesting section 23 requests the user to continue voice input, etc. “Please speak slowly. After reading the number. Please press END. "Is displayed, and an end button 126 is displayed below it.
When the user finishes pronunciation of the numbers displayed on the number display unit 122, the user presses the end button 126, thereby ending the voice input from the voice data input unit 22.
The voice data storage unit 24 temporarily stores the voice data of the user input by the voice data input unit 22 as described above (S12).

Next, the voiceprint data generation unit 28 generates voiceprint data from the voice data sent from the voice data storage unit 24 (voice data for consecutive numbers from 0 to 9) (S13).
Next, the voiceprint data registration unit 31 registers the voiceprint data (voiceprint data specified by the user ID) generated by the voiceprint data generation unit 28 in the voiceprint DB 30 (S14), and ends the process.
FIG. 13 shows a display screen of the display unit 26 from the generation of the voice print data by the voice print data generation unit 28 (S13) to the registration of the voice print data by the voice print data registration unit 31 (S14), and FIG. 14 shows the voice print data registration unit. The display screen of the display part 26 when registration of the voiceprint data by 31 is complete | finished is shown.
As shown in FIG. 13, the voiceprint data generation unit 28 generates voiceprint data from the voice data (voice data for consecutive numbers from 0 to 9) (S13), and the generated voiceprint data is a voiceprint data registration unit 31. Until the step (S14) of registering in the voiceprint DB 30, a character “under confirmation” is displayed on the character display portion 127 in the middle of the screen 120.
As shown in FIG. 14, when the voiceprint data registration by the voiceprint data registration unit 31 is completed, the character display unit 128 in the upper part of the screen 120 displays the characters “Voice registration complete”, and the middle part of the screen 120. The sentence display section 129 displays the sentence “Voice registration has been completed.”

[Authentication process by the user authentication system 20]
The authentication process by the user authentication system 20 is the same as the authentication process by the user authentication system 1.
That is, in the user authentication system 20, the voice data input unit 22, the voice data input request unit 23, the voice data storage unit 24, the character string generation unit 25, the display unit 26, the voice recognition unit 27, the voice print data generation unit 28, the voice print data. The authentication process performed by the collation unit 29 and the voiceprint DB 30 is performed in the user authentication system 1 by the voice data input unit 2, the voice data input request unit 3, the voice data storage unit 4, the character string generation unit 5, the display unit 6, and the voice recognition unit. 7. This is the same as the authentication process (steps S1 to S9 shown in FIG. 5) performed by the voiceprint data generation unit 8, the voiceprint data collation unit 9, and the voiceprint DB 10.

[Operation of User Authentication System 40]
FIG. 15 is a flowchart showing the operation of the user authentication system 40.
The operation of the user authentication system 40, that is, the authentication process in the user authentication system 40 will be specifically described below with reference to the drawings.

When the user authentication system 40 is activated and a user ID is input, first, in the terminal device 41, the character string generating unit 45 generates a random character string (S20).
The random character string generated by the character string generation unit 45 is the same as the character string generated by the character string generation unit 5 of the user authentication system 1, and here, like the character string generation unit 5, a different 5 digit number is generated each time. Generate a random number of.
Next, the random character string generated by the character string generation unit 45 is displayed on the display unit 46, and the voice data input request unit 43 requests the user to input voice (S21).
In this case, the state of the display screen of the display unit 46 is the same as the state of the display screen 110 of the display unit 6 of the user authentication system 1 (FIG. 6).
That is, on the display screen of the display unit 46, as shown in FIG. 6, the user ID “00001” is displayed on the user ID display unit 111, and the 5-digit random number generated by the character string generation unit 45 on the random number display unit 112. (Numerals) and their readings are displayed, and a text indicating that the user is requested to input speech is displayed on the text display unit 114. A start button 113 is displayed.

Next, as in the user authentication system 1, the user presses the start button 113 and reads the numbers “1”, “3”, “5”, “6”, “4” displayed on the random number display unit 112. When “1”, “san”, “go”, “roku”, “yon” are uttered by voice, the voice data input unit 42 inputs voice data issued by the user (S22).
In this case, the state of the display screen of the display unit 46 when the audio data input unit 42 is inputting audio data is the same as the state of the display screen 110 of the display unit 6 of the user authentication system 1 (FIG. 7). .
That is, on the display screen of the display unit 46, as shown in FIG. 7, the text display unit 115 displays a text to the effect that the voice data input request unit 3 requests the user to continue voice input, etc. An end button 116 and a cancel button 117 are displayed.
When the user finishes pronunciation of the number displayed on the random number display unit 112, the user presses the end button 116, thereby completing the voice input from the voice data input unit 42. Note that the user presses the cancel button 117 when canceling the voice input.
The voice data storage unit 44 temporarily stores the user voice data input by the voice data input unit 42 as described above (S23).
Next, the transmission unit 47 sends the voice data (voice data for five digits) stored in the voice data storage unit 44 and the five-digit random number (number) generated by the character string generation unit 45 to the communication line 65. To the server 50 (S24).

In the server 50, the receiving unit 51 receives voice data (voice data for a 5-digit number) and a 5-digit random number (number) transmitted from the transmitting unit 47 of the terminal device 41 (S25).
Next, the data storage unit 54 temporarily stores the voice data (voice data for a 5-digit number) and a 5-digit random number (number) sent from the receiving unit 51 (S26).

Next, the character string recognizing unit of the voice recognizing unit 57 recognizes what character string the voice data (voice data of five digits) sent from the data storage unit 54 consists of (S27).
Next, the character string determination unit of the speech recognition unit 57 performs a character string recognized by the character string recognition unit and a random character string (5-digit random number (number)) stored by the data storage unit 54 (this number is data storage). Is sent from the unit 54 to the voice recognition unit 57) (S28).
If it is determined in step S28 that they match, the determination result (successful voice recognition) is sent to the voiceprint data generation unit 58, and the process proceeds to the next voiceprint authentication step (S30). The process proceeds to step S31 without performing voiceprint authentication.

If it is determined in step S28 that they match, the voiceprint data generation unit 58 generates voiceprint data from the voice data (voice data for a 5-digit random number (number)) sent from the data storage unit 54. (S30).
Next, the voiceprint data matching unit 59 sends the voiceprint data generated by the voiceprint data generation unit 58 (this voiceprint data is sent from the voiceprint data generation unit 58 to the voiceprint data matching unit 59) and the user's voice stored in the voiceprint DB 60. The voiceprint data (voiceprint data specified by the user ID “00001”) is collated (S30).
Then, in the voiceprint data collation in S30, if the difference in the voiceprint data to be collated is small and within a certain range (below the threshold value), a collation result indicating that the voiceprint data has been successfully collated (authentication success) is output, If the difference between the voiceprint data to be collated is large and outside a certain range (greater than or equal to the threshold), a collation result indicating that the voiceprint data is unsuccessfully collated (authentication unsuccessful) is output (S31), and the process is terminated.
If it is determined in S28 that the voice data does not match, voiceprint authentication (S29, S30) is not performed, the verification result of unsuccessful authentication is output (S31), and the process ends.
The state of the display screen of the display unit 46 from the voice recognition by the voice recognition unit 57 (S27) to the output of the collation result by the voiceprint data collation unit 99 (S31) is the state of the display screen 110 of the display unit 6 of the user authentication system 1 Same as (FIG. 8).
That is, on the display screen of the display unit 46, as shown in FIG. 8, the characters “authenticating” are displayed on the character display unit 118.

  As described above, in the user authentication system 40, as in the user authentication system 1, the random character string (5-digit random number) generated by the character string generation unit 45 is different every time. The user's voice data inputted by the user is different every time, and even if the user's voice data is recorded and used for authentication, the voice recognition is unsuccessful in S28, and a verification result of unsuccessful authentication is output. Can be prevented.

[Connection to external system 100]
In the collation result output from the voiceprint data collation unit 59 in step S31, if the collation result is a successful authentication, the collation result is sent to the login unit 101 of the external system 100 and a login process is performed.
The state of the display screen of the display unit 46 after the login process is performed by the login unit 101 of the external system 100 is the same as the state of the display screen 110 of the display unit 6 of the user authentication system 1 (FIG. 9).
That is, as shown in FIG. 9, a text such as “Welcome to AA Company” is displayed on the text display unit 119 of the display screen of the display unit 46, and the user can use various services (not shown) through the external site 100. ) Can be used.

[Operation of User Authentication System 70]
As described above, since the user authentication system 70 performs voiceprint data registration processing and authentication processing in the same manner as the user authentication system 20, each of them will be described.

[Voiceprint data registration process by user authentication system 70]
FIG. 16 is a flowchart showing voiceprint data registration processing by the user authentication system 70.
The voice print data registration process by the user authentication system 70 will be specifically described below with reference to the drawings.

When the user authentication system 70 is activated and a user ID is input, first, in the terminal device 71, a predetermined character string stored in the character string storage unit 77 is displayed on the display unit 76, and a voice data input request unit 73 is displayed. Requests voice input from the user (S30).
The predetermined character string stored in the character string storage unit 77 is the same as the character string stored in the character string storage unit 21 of the user authentication system 70. Here, as in the character string storage unit 21, 0 to 9 is used. Are stored, and the display unit 76 displays the continuous numbers from 0 to 9.
In this case, the state of the display screen of the display unit 76 is the same as the state of the display screen 120 of the display unit 26 of the user authentication system 20 (FIG. 11).
That is, on the display screen of the display unit 76, as shown in FIG. 11, the character “Voice Registration” indicating that voiceprint registration is performed is displayed on the upper character display unit 121, and the character display unit 122 below the character display A continuous number from 0 to 9 stored in the column generation unit 5 is displayed together with the reading, a registration start button 123 is displayed below it, and a voice data input request unit 73 is displayed on the text display unit 124 below. A text requesting the user to input voice is displayed.

Next, similarly to the user authentication system 20, the user presses the registration start button 123, and consecutive numbers “0”, “1”, “2”, “3” displayed from 0 to 9 displayed on the number display unit 122. ,..., “9” readings “Zero”, “Ichi”, “Ni”,..., “Kyu” are uttered by voice. Is input (S31).
In this case, the state of the display screen of the display unit 76 when the audio data input unit 72 is inputting audio data is the same as the state of the display screen 120 of the display unit 26 of the user authentication system 20 (FIG. 12). .
That is, on the display screen of the display unit 76, as shown in FIG. 12, the text display unit 125 displays a text indicating that the voice data input requesting unit 73 requests the user to continue voice input, etc. An end button 126 is displayed below.
When the user finishes pronunciation of the numbers displayed on the number display unit 122, the user presses the end button 126, thereby ending the voice input from the voice data input unit 72.
The voice data storage unit 74 temporarily stores the user voice data input by the voice data input unit 72 as described above (S32).
Next, the transmission unit 78 transmits the voice data stored in the voice data storage unit 44 (voice data for consecutive numbers from 0 to 9) to the server 80 via the communication line 95 (S33).

In the server 80, the receiving unit 81 receives voice data (voice data for consecutive numbers from 0 to 9) transmitted from the transmission unit 78 of the terminal device 71 (S34).
Next, the data storage unit 84 temporarily stores the audio data sent from the receiving unit 81 (audio data for consecutive numbers from 0 to 9) (S35).
Next, the voiceprint data generation unit 88 generates voiceprint data from the voice data sent from the data storage unit 84 (voice data for consecutive numbers from 0 to 9) (S36).
Next, the voiceprint data registration unit 91 registers the voiceprint data (voiceprint data specified by the user ID) generated by the voiceprint data generation unit 88 in the voiceprint DB 90 (S37), and ends the process.
In this case, the state of the display screen of the display unit 76 when the voice print data generation unit 88 generates voice print data and the voice print data registration unit 91 registers the voice print data is the display of the display unit 26 of the user authentication system 20. It is the same as the state of the screen 120 (FIG. 13), and the state of the display screen of the display unit 76 when the voice print data registration by the voice print data registration unit 91 is completed is also the display screen 120 of the display unit 26 of the user authentication system 20. This is the same as the state (FIG. 14).

[Authentication process by the user authentication system 70]
The authentication process by the user authentication system 70 is the same as the authentication process by the user authentication system 40.
That is, in the terminal device 71 of the user authentication system 70, the voice data input unit 72, the voice data input request unit 73, the voice data storage unit 74, the character string generation unit 75, the display unit 76, the transmission unit 78, and the reception unit 79 perform. In the terminal device 41 of the user authentication system 40, the processing includes an audio data input unit 42, an audio data input request unit 43, an audio data storage unit 44, a character string generation unit 45, a display unit 46, a transmission unit 47, and a reception unit 48. It is the same as the processing to be performed.
In the server 80 of the user authentication system 70, the processing performed by the reception unit 81, the transmission unit 82, the data storage unit 84, the voice recognition unit 87, the voice print data generation unit 88, the voice print data collation unit 89, and the voice print DB 90 is user authentication. In the server 50 of the system 40, processing performed by the reception unit 51, transmission unit 52, data storage unit 54, voice recognition unit 57, voiceprint data generation unit 58, voiceprint data collation unit 59, and voiceprint DB 60 (steps S20 to S20 shown in FIG. 15). This is the same as S31).

  The user authentication system, the user authentication method, and the program of the present invention allow a user to easily input voiceprint data when performing user authentication for identity verification required by a user accessing various systems by voiceprint authentication. Voiceprint authentication with high accuracy can be performed with simple means, and voiceprint authentication voices uttered by the user can be prevented from being recorded and illegally used.

DESCRIPTION OF SYMBOLS 1 User authentication system 2 Voice data input part 3 Voice data input request part 4 Voice data storage part 5 Character string generation part 6 Display part 7 Voice recognition part 8 Voiceprint data generation part 9 Voiceprint data collation part 10 Voiceprint database (DB)
DESCRIPTION OF SYMBOLS 20 User authentication system 21 Character string storage part 22 Voice data input part 23 Voice data input request part 24 Voice data storage part 25 Character string generation part 26 Display part 27 Voice recognition part 28 Voiceprint data generation part 29 Voiceprint data collation part 30 Voiceprint database (DB)
31 Voiceprint data registration unit 40 User authentication system 41 Terminal device 42 Audio data input unit 43 Audio data input request unit 44 Audio data storage unit 45 Character string generation unit 46 Display unit 47 Transmission unit 48 Reception unit 50 Server 51 Reception unit 52 Transmission unit 54 data storage unit 57 voice recognition unit 58 voice print data generation unit 59 voice print data collation unit 60 voice print database (DB)
65, 66 Communication line 70 User authentication system 71 Terminal device 72 Audio data input unit 73 Audio data input request unit 74 Audio data storage unit 75 Character string generation unit 76 Display unit 77 Character string storage unit 78 Transmission unit 79 Reception unit 80 Server 81 Reception unit 82 Transmission unit 84 Data storage unit 87 Voice recognition unit 88 Voiceprint data generation unit 89 Voiceprint data collation unit 90 Voiceprint database (DB)
91 Voiceprint data registration unit 100 External system 101 Login unit 110 Display screen 111 User ID display unit 112 Random number display unit 113 Start button 114, 115 Text display unit 116 End button 117 Cancel button 118 Character display unit 119 Text display unit 120 Display screen,
121 Character display part 122 Number display part 123 Registration start button 124, 125 Text display part 126 End button 127, 128 Character display part 129 Text display part

Claims (13)

A character string display means for displaying a character string;
Voice data input means for inputting user voice data issued by the user;
Voice data input request means for requesting input of user voice data for the character string displayed on the character string display means;
Voiceprint data storage means for storing voiceprint data generated from user voice data issued by the user;
A character string generating means for generating a random character string;
Voice print data generating means for generating voice print data from user voice data for the random character string input from the voice data input means in accordance with a request from the voice input request means;
A user authentication system comprising voice print data collating means for collating voice print data stored in the voice print data storage means and voice print data generated by the voice print data generating means and outputting the result of the collation. A character string display means for displaying a character string;
Voice data input means for inputting user voice data issued by the user;
Voice data input request means for requesting input of user voice data for the character string displayed on the character string display means;
Character string storage means for storing a predetermined character string;
A character string generating means for generating a random character string;
Voice print data generation means for generating voice print data from user voice data for the predetermined character string and the random character string input by the voice data input means in accordance with a request from the voice input request means;
Voiceprint data registration means for registering voiceprint data generated from user voice data for the predetermined character string;
Voiceprint data collating means for collating voiceprint data generated from user voice data for the random character string with voiceprint data registered by the voiceprint data registering means and outputting the collation result And a user authentication system.   In accordance with a request from the voice input request means, a character string is recognized from user voice data for the random character string input from the voice data input means, and the recognized character string and the random character string are The user authentication system according to claim 1 or 2, further comprising voice recognition means for determining whether or not they match. A character string display means for displaying a character string;
Voice data input means for inputting user voice data issued by the user;
Voice data input request means for requesting input of user voice data for the character string displayed on the character string display means;
A character string generating means for generating a random character string;
A terminal device comprising user voice data transmitting means for transmitting user voice data for the random character string input from the voice data input means in accordance with a request from the voice input request means;
User voice data receiving means for receiving user voice data for the random character string transmitted by the user voice data transmitting means;
Voiceprint data storage means for storing voiceprint data generated from user voice data issued by the user;
Voiceprint data generating means for generating voiceprint data from user voice data for the random character string received by the user voice data receiving means;
A server including voice print data stored in the voice print data storage means and voice print data generated by the voice print data generation means, and a voice print data check means for outputting the result of the check; User authentication system. A character string display means for displaying a character string;
Voice data input means for inputting user voice data issued by the user;
Voice data input request means for requesting input of user voice data for the character string displayed on the character string display means;
Character string storage means for storing a predetermined character string;
A character string generating means for generating a random character string;
A terminal device comprising user voice data transmission means for transmitting user voice data for the predetermined character string and the random character string input from the voice data input means in accordance with a request from the voice input request means. When,
User voice data receiving means for receiving user voice data for the predetermined character string and the random character string transmitted by the user voice data transmitting means;
Voiceprint data generating means for generating voiceprint data from user voice data for the predetermined character string and the random character string received by the user voice data receiving means;
Voiceprint data registration means for registering voiceprint data generated from user voice data for the predetermined character string by the voiceprint data generation means;
Voiceprint data collating means for collating voiceprint data generated from user voice data for the random character string by the voiceprint data generating means with voiceprint data registered by the voiceprint data registering means and outputting the collation result; A user authentication system comprising: a server equipped with a server.   The server recognizes a character string from user voice data regarding the random character string received by the user voice data receiving means, and determines whether the recognized character string matches the random character string. 6. The user authentication system according to claim 4, further comprising voice recognition means. Display the random character string generated by the character string generation means on the character string display means,
According to the request from the voice data input request means, user voice data issued by the user for the random character string displayed on the character string display means is input from the voice data input means,
Voiceprint data generating means extracts voiceprint data from user voice data for the random character string input from the voice data input means;
Voice print data generated from user voice data issued by the user in advance is stored in the voice print data storage means,
User authentication comprising: a step of collating voiceprint data stored in the voiceprint data storage means with voiceprint data extracted by the voiceprint data generating means and outputting a result of the collation Method. A predetermined character string stored in the character string storage means is displayed on the character string display means;
In accordance with a request from the voice data input request means, user voice data issued by the user for the predetermined character string displayed on the character string display means is input from the voice data input means,
Voiceprint data generating means generates voiceprint data from user voice data for the predetermined character string input from the voice data input means;
Voiceprint data registration means registers voiceprint data generated from user voice data for the predetermined character string by voiceprint data generation means;
Display the random character string generated by the character string generation means on the character string display means,
According to a request from the voice data input request means, user voice data issued by the user for the random character string displayed on the character string display means is input from the voice data input means,
The voiceprint data generating means generates voiceprint data from user voice data for the random character string input from the voice data input means;
Voiceprint data collating means collates the voiceprint data generated from the user voice data for the random character string by the voiceprint data generating means and the voiceprint data registered by the voiceprint data registering means, and outputs the collation result A user authentication method comprising the step of:   A voice recognition unit recognizes a character string from user voice data for the random character string input from the voice data input unit in accordance with a request from the voice input request unit, and the recognized character string and the 9. The user authentication method according to claim 7, further comprising a step of determining whether or not the artificial character string matches. Display the random character string generated by the character string generation means on the character string display means,
According to the request from the voice data input request means, user voice data issued by the user for the random character string displayed on the character string display means is input from the voice data input means,
User voice data transmission means transmits user voice data for the random character string input from the voice data input means,
User voice data receiving means receives user voice data for the random character string transmitted by the user voice data transmitting means,
Voiceprint data generating means extracts voiceprint data from user voice data for the random character string received by the user voice data receiving means,
Voice print data generated from user voice data issued by the user in advance is stored in the voice print data storage means,
A user authentication comprising: a step of collating voiceprint data stored in the voiceprint data storage unit with voiceprint data generated by the voiceprint data generation unit and outputting a result of the collation Method. A predetermined character string stored in the character string storage means is displayed on the character string display means;
In accordance with a request from the voice data input request means, user voice data issued by the user for the predetermined character string displayed on the character string display means is input from the voice data input means,
User voice data transmission means transmits user voice data for the predetermined character string input from the voice data input means;
User voice data receiving means receives user voice data for the predetermined character string transmitted by the user voice data transmitting means;
Voice print data generating means generates voice print data from user voice data for the predetermined character string received by the user voice data receiving means;
Voiceprint data registration means registers voiceprint data generated from user voice data for the predetermined character string by voiceprint data generation means;
Display the random character string generated by the character string generation means on the character string display means,
According to a request from the voice data input request means, user voice data issued by the user for the random character string displayed on the character string display means is input from the voice data input means,
The user voice data transmitting means transmits user voice data about the random character string input from the voice data input means;
The user voice data receiving means receives user voice data for the random character string transmitted by the user voice data transmitting means;
The voiceprint data generating means generates voiceprint data from user voice data for the random character string received by the user voice data receiving means;
Voiceprint data collating means collates the voiceprint data generated from the user voice data for the random character string by the voiceprint data generating means and the voiceprint data registered by the voiceprint data registering means, and outputs the collation result A user authentication method comprising the step of:   Voice recognition means recognizes a character string from user voice data for the random character string received by the user voice data receiving means, and whether the recognized character string matches the random character string. The user authentication method according to claim 10 or 11, further comprising a determination step. A program for causing a computer to execute the user authentication method according to any one of claims 7 to 12.

Images