JP6370172B2 - Communication device, voiceprint registration method and program - Google Patents

Description

  The present invention relates to a call device, a voiceprint registration method, and a program for acquiring and automatically registering voiceprint data for voiceprint authentication in real time from a received voice during a call without being aware of the voice of the other party.

  In recent years, nuisance calls such as “Ole scam” phone, sales phone call or mischief call pretending to be the principal have become a social problem. As a countermeasure against annoying calls by such a malicious third party, a system for authenticating a caller using a voiceprint is disclosed.

  For example, in a voice print registration database storing voice print data corresponding to a telephone, and in a telephone network of call connection established between the first telephone and the second telephone established by a call from the first telephone Route detection means for detecting a route, voice acquisition means for acquiring voice data transmitted from the first telephone to the second telephone from a path detected by the path detection means, and voice pattern registration database Referring to the stored voiceprint data corresponding to the second telephone set, voiceprint matching means for matching the voiceprint of the voiceprint data with the voiceprint of the voice data acquired by the voice acquisition means Voice print data is acquired from the voice of the sender sent to the receiver via the telephone network, and the collation result of the acquired voice print data and the voice print data stored in the voice print database is sent to the receiver. Subjecting technique is disclosed (for example, see Patent Document 1).

  In addition, a voice print storage unit storing collation voice print data, a voice acquisition unit for acquiring the voice of the communication partner, and extracting voice print data from the voice acquired by the voice acquisition unit, and collating with the voice print data for verification There is also disclosed a technique that includes a voiceprint matching unit that performs a voiceprint data registration operation, and includes a result notification unit that notifies a matching result of the voiceprint matching unit (see, for example, Patent Document 2).

JP 2006-38955 A JP 2013-164472 A

  However, in the technique described in Patent Document 1, voiceprint data is registered by accessing the system via the Web, clearing user authentication, and then calling the other party who wants to register and storing the voiceprint data of the other party in the system. There is a problem that the operation is troublesome.

  Moreover, although the technique described in Patent Document 2 can eliminate the cumbersome operation in terms of words, the criteria for determining the conditions for acquiring a voiceprint and in what format is used. There is no specific presentation such as whether to determine or at what timing to notify the result of voiceprint authentication, and it is unclear how to implement the technology, and there is a problem of lack of practicality.

  Furthermore, in the past, customer support centers (call centers) and financial institutions have been listening to pre-set passwords from customers over the phone and authenticating voiceprints, but this method checks for voiceprint matches. Therefore, it is necessary to set in advance a fixed passphrase that makes it easy for the customer to remember the “telephone number”, “account number”, etc., and there is a problem that the serviceability is poor. In addition, in order to register a customer's voiceprint, an operation by an operator such as a call center is required, and there is a problem that costs increase.

  Therefore, the present invention has been made in view of the above-described problems, and automatically acquires voiceprint data for voiceprint authentication from a received voice during a call in real time without being aware of the voice of the other party. An object of the present invention is to provide a call device, a voiceprint registration method, and a program for registration in the Internet.

  (1) The present invention provides a telephone book, voice acquisition means for acquiring a received voice, telephone number acquisition means for acquiring a telephone number on the utterance side, extraction means for extracting voiceprint data from the acquired received voice, Measuring means for measuring the acquisition time of the received voice; and storage means for storing the acquired telephone number and the voiceprint data extracted by the extracting means in association with each other when the acquired telephone number is in the telephone directory; Determining whether or not the total acquisition time length of at least one voiceprint data corresponding to the same telephone number as the acquired telephone number is longer than a time required for voiceprint matching in the telephone directory However, in the case where the telephone number is long, there is proposed a communication device including a control unit that stores the acquired telephone number and voiceprint data in association with each other in the storage unit.

  (2) The present invention provides a volume acquisition unit that acquires the volume of the received voice and a comparison unit that compares the volume of the received voice with a predetermined volume threshold for the call device according to (1). And the measuring means measures a time when the volume of the received voice exceeds a predetermined volume threshold in the comparing means.

  (3) In the communication device of (1), when the extracted voiceprint data is shorter than the time required for voiceprint matching, the present invention determines whether or not to store the voiceprint data in the storage means. Or a communication device comprising a selection means for the user to select whether or not to store the voiceprint data when the extracted voiceprint data is longer than the time required for voiceprint matching. Has proposed.

  (4) The present invention is a voiceprint registration method in a telephone conversation device comprising voice acquisition means, telephone number acquisition means, extraction means, measurement means, storage means, and control means, wherein the voice acquisition means A first step of acquiring, a second step of acquiring the telephone number of the utterance by the telephone number acquiring means, and a third step of extracting voice print data from the acquired received voice by the extracting means; A fourth step in which the measuring means measures the acquisition time of the received voice; and at least one voiceprint corresponding to the same telephone number as the acquired telephone number, wherein the control means is in a telephone directory. It is determined whether or not the total acquisition time length of the data is longer than the time required for voiceprint collation. If it is longer, the acquired telephone number and voiceprint data are stored in the storage means in association with each other. 5th Proposes a voiceprint registration method characterized by comprising the steps, a.

  (5) The present invention is a program for causing a computer to execute a voiceprint registration method in a telephone conversation device comprising voice acquisition means, telephone number acquisition means, extraction means, measurement means, storage means, and control means, A first step in which the voice acquisition means acquires the received voice; a second step in which the telephone number acquisition means acquires the telephone number on the utterance side; and A third step of extracting the received voice, a fourth step in which the measuring means measures the acquisition time of the received voice, and the control means is in the telephone directory, and the same telephone number as the acquired telephone number. It is determined whether or not the combined acquisition time length of at least one or more corresponding voiceprint data is longer than the time required for voiceprint collation, and if so, the acquired telephone number and voiceprint data are acquired. We propose a program for executing a fifth step of storing in the storage means in association bets, to the computer.

  According to the present invention, it is possible to automatically register a voiceprint of a call partner during a call, and there is an effect that it is not necessary for the user to perform complicated operations.

It is a figure which shows the structure of the telephone apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the detailed structure of the volume acquisition part of the telephone apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the relationship between the buffer maximum value which concerns on the 1st Embodiment of this invention, and time. It is a figure which shows the relationship between the utterance time which concerns on the 1st Embodiment of this invention, and a silence time. It is a figure which shows the data structure at the time of the voiceprint registration which concerns on the 1st Embodiment of this invention. It is a figure which shows the registration process of the voiceprint at the time of the telephone number transmission / reception already registered into the telephone directory which concerns on the 1st Embodiment of this invention. It is a figure which shows the audio | voice recording process of the other party who concerns on the 1st Embodiment of this invention. It is a figure which shows the registration process of the voiceprint at the time of the telephone number transmission / reception unregistered in the telephone directory which concerns on the 1st Embodiment of this invention. It is a figure which shows the structure of the telephone apparatus which concerns on the 2nd Embodiment of this invention. It is a figure which shows the process of the telephone apparatus which concerns on the 2nd Embodiment of this invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
Note that the constituent elements in the present embodiment can be appropriately replaced with existing constituent elements and the like, and various variations including combinations with other existing constituent elements are possible. Therefore, the description of the present embodiment does not limit the contents of the invention described in the claims.

<First Embodiment>
Hereinafter, a call device according to a first embodiment of the present invention will be described with reference to FIGS.

<Configuration of telephone device>
As shown in FIG. 1, the communication device 100 according to the present embodiment includes a voice acquisition unit 110, an extraction unit 120, a volume acquisition unit 130, a comparison unit 140, an utterance time acquisition unit 141, and a measurement unit 150. , Control unit 160, telephone number acquisition unit 170, telephone book 180, and storage unit 190.

  The voice acquisition unit 110 acquires the received voice from the speaker. The telephone number acquisition unit 170 acquires the telephone number on the utterance side. By comparing the acquired telephone number with the telephone number registered in the telephone directory 180, it can be determined whether or not the utterance number is a number registered in the telephone directory. This determination is processed in the control unit 160.

  The extraction unit 120 extracts voiceprint data from the received voice acquired by the voice acquisition unit 110. The human voice acquired by the voice acquisition unit 110 is composed of a collection of sound waves of various frequencies. The extraction unit 120 extracts a voiceprint by graphing the received voice acquired from the voice acquisition unit 110 with parameters of time, frequency, and intensity. Here, the extracted voiceprint has a characteristic of drawing different characteristic patterns for each person like a fingerprint. In this specification, “voice print” in the case of “voice print”, “voice print data”, “voice print verification”, and “voice print registration” indicates what has been described above, and in the case of “voice” Refers to “a human voice composed of a collection of sound waves of various frequencies”.

  The measurement unit 150 measures the acquisition time of the received voice acquired by the voice acquisition unit 110. The volume acquisition unit 130 acquires the volume from the received voice acquired by the voice acquisition unit 110. The comparison unit 140 compares the volume of the received voice acquired by the volume acquisition unit 130 with a predetermined volume threshold. The utterance time acquisition unit 141 acquires the utterance time from the received data. That is, the measuring unit 150 uses the comparison unit 140 to measure from the acquired data of the utterance time acquiring unit 141 the time when the volume of the received voice acquired by the volume acquiring unit 130 exceeds a predetermined volume threshold. Therefore, it is possible to determine whether or not this integrated value satisfies the time required for voiceprint matching by integrating the time exceeding the threshold as the utterance time.

  When the telephone number acquired by the telephone number acquisition unit 170 is a telephone number registered in the telephone directory 180, the storage unit 190 stores the telephone number and the voiceprint data extracted by the extraction unit 120 in association with each other.

  The control unit 160 controls the entire call device according to a control program recorded in a ROM (Read Only Memory). For example, the control unit 160 determines whether or not the utterance time is sufficient for voiceprint matching. Further, when the utterance time is sufficient for voiceprint collation and the acquired telephone number is registered in the telephone directory, the telephone number and voiceprint data are stored in association with each other in response to a registration request from the user. . Further, when the utterance time is not sufficient for voiceprint collation and the acquired telephone number is registered in the telephone directory, the voiceprint data is accumulated and stored in response to a request from the user.

  Although not shown, a selection unit may be provided for the user to select whether the extracted voiceprint data is registered in the storage unit 190 or stored and stored.

<Detailed configuration of volume acquisition unit>
As shown in FIG. 2, the volume acquisition unit 130 includes a received voice downlink buffer 131, a buffer maximum value acquisition unit 132, and an audio file database 133.

  The received voice downlink buffer 131 samples and stores the voice data acquired by the voice acquisition unit 110. For example, the specifications of the received voice downlink buffer 131 are a sampling rate of 16 kHz, a buffer size of 8 kB, and a buffer time of 0.5 sec.

The buffer maximum value acquisition unit 132 acquires the maximum value of the data sampled within the buffer time. The voice file database 133 stores the data when the buffer in the received voice downlink buffer 131 expires.
The output of the volume acquisition unit 130 is as shown in FIG.

  Then, the output of the sound volume acquisition unit 130 is passed to the comparison unit 140 and compared with a predetermined threshold value shown in FIG. 3, and based on the result, the measurement unit 150 separates the call time into speech time and silent time. And measure the total utterance time.

<Voiceprint registration process for incoming / outgoing phone numbers registered in the phone book>
The voiceprint registration process at the time of incoming / outgoing telephone numbers registered in the telephone directory according to the present embodiment will be described with reference to FIGS.

  As shown in FIG. 6, when a voice call is received from a telephone number registered in the telephone directory and a voice call is made to the telephone number, the voice pattern of the other party is registered according to the following procedure.

  When there is an incoming call, recording of the received voice starts automatically. Then, the utterance time of the calling party excluding the silent gap is regularly observed (step S101). At this time, it is determined whether or not a valid utterance time has been recorded (step S102). If a valid utterance time has been recorded ("YES" in step S102), the previously accumulated voice is deleted. (Step S103).

  Next, a selection screen showing a message as to whether or not to register a voiceprint is displayed (step S104), and when the user selects “register”, as shown in FIG. A voiceprint file is generated from the phonebook, and registered in association with the telephone number of the telephone directory (step S105). On the other hand, when the user selects “do not register”, the received received voice is discarded (step S106).

  If it is not possible to record a valid utterance time in step S102 ("NO" in step S102), it is confirmed whether or not the call is ended (step S109). If the call is not ended (step S109). If "NO"), the process returns to step S101.

  On the other hand, if it is determined that the conversation is over (“YES” in step S109), a selection screen showing a message as to whether or not to store voiceprints is displayed (step S110), and the user “does not store” "Is selected, the recorded received voice is discarded (step S111).

  If the user selects “accumulate”, the recorded received voice is accumulated and stored (step S112). Then, the received received voice is stored in the database, and when the recorded time of the received voice is a valid recording time (“YES” in step S107), the process proceeds to step S104. On the other hand, if the recording time of the accumulated received voice is not an effective recording time (“NO” in step S107), no special processing is performed (step S108).

<Call voice recording process>
A call voice recording process according to the present embodiment will be described with reference to FIG.

  When the telephone call is started (step S201), the reception voice recording is started and the utterance time is measured (step S202). Next, effective voicing time measurement is started (step S203).

  Specifically, a check timer is started (step S204), and it is determined whether the timer has expired (step S205). If the timer has not expired (“NO” in step S205), the process returns to the original state. If the timer has expired (“YES” in step S205), the utterance time is then calculated (step S206). It is determined whether or not the utterance time is longer than 60 seconds (step S207).

  If the utterance time is shorter than 60 seconds (“NO” in step S207), the process returns to step S204. If the utterance time is longer than 60 seconds (“YES” in step S207), the received voice is End recording.

<Voiceprint registration process for incoming / outgoing phone numbers not registered in the phone book>
With reference to FIG. 8, a voiceprint registration process at the time of incoming / outgoing telephone numbers not registered in the telephone directory according to the present embodiment will be described.

  When there is an incoming call, recording of the received voice starts automatically. Then, the utterance time of the calling party excluding the silent gap is regularly observed (step S101). At this time, it is determined whether or not a valid utterance time has been recorded (step S102). If a valid voicing time has been recorded ("YES" in step S102), whether or not to register a voiceprint is determined. A selection screen showing such a message is displayed (step S301), and when the user selects “register”, a voice print file is generated from the recorded received voice (step S302). On the other hand, when the user selects “do not register”, the recorded received voice is discarded (step S303).

  If it is not possible to record a valid utterance time in step S102 ("NO" in step S102), it is confirmed whether or not the call is ended (step S109). If the call is not ended (step S109). If "NO"), the process returns to step S101.

  On the other hand, if it is determined that the conversation is over ("YES" in step S109), a selection screen showing a message indicating that sufficient voice necessary for voiceprint registration could not be collected is displayed (step S304). When the user selects “OK”, the received received voice is discarded (step S305).

  As described above, according to the present embodiment, it is possible to automatically register the voiceprint of the other party during a call, so there is no need for the user to perform complicated operations.

<Second Embodiment>
The second embodiment according to the present invention will be described below with reference to FIGS. 9 and 10.

<Configuration of telephone device>
As shown in FIG. 9, the communication device 100 according to the present embodiment includes a voice acquisition unit 110, an extraction unit 120, a volume acquisition unit 130, a comparison unit 140, an utterance time acquisition unit 141, and a measurement unit 150. The control unit 161, the telephone number acquisition unit 170, the phone book 180, and the storage unit 190 are configured. In addition, about the component which attaches | subjects the same code | symbol as 1st Embodiment, since it has the same function, the detailed description is abbreviate | omitted.

  The control unit 161 has the acquired telephone number registered in the telephone directory, and the combined acquisition time length of at least one voiceprint data corresponding to the same telephone number as the acquired telephone number is necessary for voiceprint matching. It is determined whether or not the time is longer than the predetermined time. If the time is longer, the acquired telephone number and voiceprint data are stored in the storage unit 190 in association with each other.

<Handling device processing>
The processing of the call device according to this embodiment will be described with reference to FIG.

  First, the voice acquisition unit 110 acquires the received voice (step S410), the telephone number acquisition unit 170 acquires the telephone number on the utterance side (step S420), and the extraction unit 120 extracts voiceprint data from the acquired received voice. Is extracted (step S430).

  Next, the measurement unit 150 measures the acquisition time of the received voice (step S440). Whether the combined acquisition time length of at least one or more voiceprint data corresponding to the same telephone number as the acquired telephone number is longer than the time required for voiceprint matching in the control unit 161 If it is long, the acquired telephone number and voiceprint data are stored in the storage unit 190 in association with each other (step S450).

  As described above, according to the present embodiment, it is possible to automatically register the voiceprint of the other party during a call, so there is no need for the user to perform complicated operations.

  It is to be noted that the call device of the present invention can be realized by recording the processing of the call device on a recording medium readable by the computer system, causing the call device to read and execute the program recorded on the recording medium. The computer system here includes an OS and hardware such as peripheral devices.

  Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW (World Wide Web) system is used. The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.

  The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  The embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to the embodiments, and includes designs and the like that do not depart from the gist of the present invention. For example, a system may be provided in which a server is provided and the functions of the extraction unit 120, the volume acquisition unit 130, the comparison unit 140, the measurement unit 150, the control unit 161, and the telephone number acquisition unit 170 among the functions of the communication device are realized by the server. .

  Further, by loading the program of the present invention, the present invention can be realized not only in a mobile phone and a smart phone but also in a fixed phone connected by a telephone line.

DESCRIPTION OF SYMBOLS 100; Telephone apparatus 110; Voice acquisition part 120; Extraction part 130; Volume acquisition part 131; Received voice downlink buffer 132; Buffer maximum value acquisition part 133; Voice file database 141; Speech time acquisition part 140; Unit 160; control unit 161; control unit 170; telephone number acquisition unit 180; telephone directory 190; storage unit

Claims (5)

Phone book and
Voice acquisition means for acquiring the received voice;
A telephone number acquisition means for acquiring the telephone number of the utterance;
Extraction means for extracting voice print data from the acquired received voice;
Measuring means for measuring the acquisition time of the received voice;
Storage means for storing the acquired telephone number and the voiceprint data extracted by the extracting means in association with each other when the acquired telephone number is in the telephone directory;
It is determined whether the total acquisition time length of at least one voiceprint data corresponding to the same telephone number as the acquired telephone number is longer than a time required for voiceprint matching in the telephone directory. If long, the control means for associating the acquired telephone number and voiceprint data to be stored in the storage means,
A call device characterized by comprising: Volume acquisition means for acquiring the volume of the received voice;
Comparing means for comparing the volume of the received voice with a predetermined volume threshold;
With
The call device according to claim 1, wherein the measuring means measures a time when the volume of the received voice exceeds a predetermined volume threshold in the comparing means.   If the extracted voiceprint data is shorter than the time required for voiceprint matching, whether or not the voiceprint data is accumulated and stored in the storage means, or the extracted voiceprint data is used for voiceprint matching. 2. The call device according to claim 1, further comprising a selection unit that allows a user to select whether or not to store the voiceprint data when the time is longer than a necessary time. A voiceprint registration method in a communication device comprising voice acquisition means, telephone number acquisition means, extraction means, measurement means, storage means, and control means,
A first step in which the voice acquisition means acquires a received voice;
A second step in which the telephone number acquisition means acquires the telephone number of the utterance;
A third step in which the extraction means extracts voiceprint data from the acquired received voice;
A fourth step in which the measuring means measures an acquisition time of the received voice;
Whether the control means is in the telephone directory and the total acquisition time length of at least one voiceprint data corresponding to the same telephone number as the acquired telephone number is longer than the time required for voiceprint matching And if it is longer, a fifth step of storing the acquired telephone number and voiceprint data in association with each other in the storage means;
A voiceprint registration method characterized by comprising: A program for causing a computer to execute a voiceprint registration method in a call device provided with voice acquisition means, telephone number acquisition means, extraction means, measurement means, storage means, and control means,
A first step in which the voice acquisition means acquires a received voice;
A second step in which the telephone number acquisition means acquires the telephone number of the utterance;
A third step in which the extraction means extracts voiceprint data from the acquired received voice;
A fourth step in which the measuring means measures an acquisition time of the received voice;
Whether the control means is in the telephone directory and the total acquisition time length of at least one voiceprint data corresponding to the same telephone number as the acquired telephone number is longer than the time required for voiceprint matching And if it is longer, a fifth step of storing the acquired telephone number and voiceprint data in association with each other in the storage means;
A program that causes a computer to execute.