JP4812928B2 - Voiceprint phone system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a voiceprint telephone system, and more particularly to a voiceprint telephone system that recognizes specific data from a voiceprint of voice data from a telephone and enables connection to the other party.
[0002]
[Prior art]
Telephones in the prior art, especially mobile phones, are rapidly becoming popular, and they add value in addition to this conversation because they use a conventional numeric keypad and compression function to call a specific destination and have a conversation. There are various ways to use such services.
[0003]
[Problems to be solved by the invention]
However, in the above-described telephone set, particularly a mobile phone, a voice recognition telephone that recognizes voice is already well known. However, since the telephone or the terminal that receives voice is recognized as a real problem, the voice recognition rate is low. There is a problem of being bad. When the mobile phone is lost, it is necessary to re-create the telephone directory registered in the lost mobile phone, and it is necessary to change the phone number every time the phone number is changed. Furthermore, there is no problem with using only the phone, but it is necessary to create functions other than the numeric keypad for operations such as registration of phone numbers and reservations. Therefore, the current situation is that it is advertised but not actually used. In addition, since the operation surface of a mobile phone is generally small, the letters and numbers to be displayed are also small, and it has been pointed out that it is more difficult to see when used in a person with low vision or in a dark place.
[0004]
Therefore, there is a problem that a function derived from this method must be solved based on a method in which a telephone can call and talk to a desired partner without using a numeric keypad.
[0005]
[Means for Solving the Problems]
In order to solve the above problems, the voiceprint telephone system according to the present invention is configured as follows.
[0006]
(1) A telephone having at least a microphone, a speaker, and power means, and the telephone calls a voiceprint telephone call center having a function of decoding a predetermined voiceprint when the power means can make a call. The voice data inputted from the microphone is decoded at the voice print telephone call center side, and when the decoded voice print is recognized as the specific voice data, the voice data is stored. A voiceprint telephone system having a function of calling a destination based on a stored destination number corresponding to the stored voice data if it is stored as a result of the search .
[0007]
As described above, a call center is provided between the telephone and a specific voice data is recognized by a voiceprint so that a conversation or the like can be performed with the other party. By storing the number of the other party once used as voice data, for example, it becomes possible to connect to the other party in a normal conversation format such as “Mr. A who went yesterday” or “B-kun”.
[0008]
DETAILED DESCRIPTION OF THE INVENTION
Next, various embodiments of a voiceprint telephone system according to the present invention will be described with reference to the drawings.
[0009]
As shown in FIG. 1, the voiceprint telephone system according to the first embodiment of the present invention is a system in which a voiceprint telephone call center 13 can be called from a voiceprint telephone 10 or a numeric keypad telephone 11 to make a telephone call to a desired partner. Yes, the voice print type telephone 10 is connected to the voice print telephone call center 13 automatically or by voice, and the ten key type telephone is connected to the voice print telephone call center 13 by pushing the numeric keypad.
[0010]
The voice print telephone call center 13 is connected to the voice print type or numeric keypad type telephones 10, 11. When voice data 14 is input from any of the telephones 10, 11, the voice print 14 of the voice data 14 is decoded 15. If it matches the preset specific voice print here, it means that the person who can use the voice print telephone call center has been authenticated, and the words contained in the voice data 14 have been translated and decoded in advance. The meaning of the wording is deciphered by associating it with the voiceprint database that stores the data.
[0011]
Such a translation function 16 includes a quantitative learning master function and a qualitative learning master function. The quantitative learning master function simply accumulates (accumulates) information, which is like an encyclopedia. It is necessary to have a mechanism that automatically propagates as a volume of information. In fact, as new words continue to appear, they are registered.
[0012]
The qualitative learning master function, which should be called artificial intelligence, is the center of the translation function 16, and the characteristic point is that it focuses on the habits of people and companies. In other words, if we focus on the fact that the qualitative ability of translation depends on customs, customs (customs) are based on merit. Therefore, if there is no merit, it will not be established as a habit. For example, there is a custom that individuals with whom an individual calls are almost specified and limited. In addition, greetings and New Year's cards are almost standardized and customary. Furthermore, companies are also specified and limited by business practices such as trading practices. In the retail industry, there is a business practice that A manufacturer's products are delivered via B wholesaler. Therefore, if only the manufacturer name is determined when a new product is released, the name of the transaction wholesaler can be automatically determined. Similarly, it can be applied to genre discrimination of new products. Sales margins are also customary, and by using this custom learning master, the retail sales price can be automatically determined. In this way, both individuals and companies are governed by customs and are acting according to habits, and learning habits improves their translation skills.
[0013]
Using such a translation function, a desired item is determined from the voice data 14. In the embodiment, if it is determined that “Mr. A of the company oo” is determined, the process goes to the destination connection routine 17; if “E-mail” is determined, the process goes to the voice-email routine 18; Go to the word processor routine 19, go to the “voice internet” routine 20 if it is “Internet”, and go to the voice electronic commerce routine 21 if it is “electronic commerce”. Hereinafter, the configuration and processing form of each routine will be described.
[0014]
As shown in FIG. 2, the configuration in the routine for connecting to the other party and the processing form thereof can be connected to the network 31 including a voiceprint telephone 30, an exchange network, etc. that can be connected to the telephone, and the network 31. A voice print telephone call center 13A that can be made and a telephone 32 of an unspecified number of other parties. The network 31 is not limited as long as the telephones can be connected to each other, and is a concept that includes everything from a home-use to an office-use corporate LAN or the like, or a telephone exchange network and a satellite network.
[0015]
As shown in FIG. 2B, the voiceprint phone 30 is a so-called mobile phone, and includes a connection unit 36 that makes a connection request to at least the microphone 33, speaker 34, power means 35, and network 31. And a data transmission / reception unit 37 that processes the audio signal input from the microphone 33 and transmits the audio data to the network 31 or receives the audio data via the network 31. Here, the voice print telephone call center 13A is automatically called and connected to the connection request signal for the network 31. The power means 35 is composed of a switch that is manually turned on. When the power means 35 is turned on, a connection request signal is automatically transmitted from the connection unit 36 and is connected to the voiceprint telephone call center 13A. The switch as the power means 35 may be turned on based on an audio signal input from the microphone 33. Further, when the voice print telephone call center 13A is connected, a specific display means indicating that voice input is possible may be turned on or blinked so that it can be visually recognized. The data transmission / reception unit 37 has a function of recording specific voiceprint data in advance. When voice other than the specific voiceprint is input, the voice transmission / reception unit 37 deletes the voice data other than the voiceprint data to the voiceprint telephone call center 13A. Acts as a sending filter. Therefore, it is possible to send only a specific person's voice without being influenced by the external environment using the voiceprint telephone 30. In addition, this filter may have a function that is not used by anyone other than a specific person.
[0016]
As shown in FIG. 2A, the voiceprint telephone call center 13A is a center that can respond to connection requests from an unspecified number of voiceprint telephones 30, and is connected to the telephone 32 on the opposite side requested from the voiceprint telephone 30. Switching unit 40, voice data receiving unit 41 for receiving voice data from the voiceprint telephone 30, and voiceprint analyzing unit 42 for recognizing voice data of the received voice data and recognizing that it is specific voice data A voice print search unit 43 for searching for a telephone number or the like connected to the other party based on the specific voice data obtained by the voice print analysis unit 42, and a voice print search unit 43 A voiceprint database 44 in which data such as connection requests by a specific person having specific audio data is stored, and the other party based on the searched data And a destination connection portion 45 that connects to the phone.
[0017]
As shown in FIG. 3, a general processing form in the voiceprint telephone system having such a configuration is as follows. When the power of the voiceprint telephone 30 owned by “Mr. The calling operation for calling is performed (step ST10).
[0018]
When the voice print phone 30 and the voice print phone call center 13A are in a connected state, if voice for calling the other party is generated from the voice print phone 30, the voice data is received by the voice print phone call center 13A, and the voice print is identified and decoded. (Steps ST11 and ST12).
[0019]
If the voiceprint decoding is successful, the database is searched based on the data for identifying the other party included in the voice data, for example, “Mr. A”, which has been used in the past or entered in advance. Search for data of a certain “Mr. A”. When a hit is made, the telephone of “Mr. A” who is the other party is called (steps ST13 and ST14).
[0020]
The act of calling the other party's “Mr. A” phone starts, and by switching the connection between the caller and the callee, the caller “Mr. Ko” connects to the caller “Mr. A” and has a conversation, etc. (Steps ST15 and ST16).
[0021]
Now, only in such a general connection state, the caller side and the callee side cannot be connected, and various processes are necessary for connection. In the following, with reference to the overall configuration diagram shown in FIG. 2, the description will be divided into calling from the calling side, voiceprint decoding, search by voiceprint database, and the like.
[0022]
Invoking the voiceprint telephone call center 13A from the calling party on the voiceprint telephone 30 side is conditional on the voiceprint telephone 30 being switched on as shown in FIG. 4 (steps ST20 and ST21). That is, when inputting the voice of the caller side, it is necessary to be connected to the voiceprint telephone call center 13A in advance. This connection may be activated by turning on a mechanical power switch, or by inputting any voice signal on the caller side by software or identifying a specific voice signal to the voiceprint telephone call center 13A. You may make it connect.
[0023]
Thus, when the switch is turned on, automatic calling is performed by a specific telephone number for calling the voiceprint telephone call center 13A (step ST22). This specific telephone number is incorporated in advance in the voiceprint telephone 30 and is composed of a single telephone number or a plurality of telephone numbers.
[0024]
An automatic call is made, a connection is attempted, and when the connection is successfully completed, a voice input timer is started and a voice input from the microphone is waited. If there is no voice input within a predetermined time after the timer starts, the timer will time out and turn off the power switch (ST23, ST24, ST25, ST26). When an audio signal is input within the timer time, the input audio data is sent to the voiceprint telephone call center 13A, where voiceprint recognition, connection to a desired destination, and the like are performed (step ST26). This point will be described in detail in the voice print telephone call center routine described later.
[0025]
In step ST23, if the connection by automatic call fails, the recall counter is incremented, and if it is less than the preset count value, automatic call retry is attempted again (steps ST27, ST28, ST29). ). If the value of the recall count reaches a preset value or more, the first telephone number by automatic call is terminated, and the automatic call is continued by switching to the next second telephone number (step). ST30, ST31). Here, if the connection fails even if all of the automatic call telephone numbers are used, a manual voice number calling routine is performed in order to establish a manual connection.
[0026]
FIG. 5 shows a manual voice number calling routine, which manually calls the voiceprint telephone call center 13A. In order to call this voice print telephone call center 13A, voice input (encrypted code voice data) of a caller-specific recitation number is required (step ST40).
[0027]
The voiceprint telephone 30 performs signal processing on the personal identification number, converts it into encrypted code voice data that is a telephone number for calling the voiceprint telephone call center 13A, and transmits it to call the voiceprint telephone call center (steps ST41 and ST42).
[0028]
When the connection is completed, the voice input timer shown in FIG. 4 is started and the voice input on the caller side is awaited (step ST43).
[0029]
In step ST43, if the connection fails and the recall count value is equal to or smaller than the predetermined value, the voiceprint telephone call center is called again to retry (steps ST44, ST45, ST46). In step ST43, if the connection fails, the recall counter is incremented, and if the count value is equal to or greater than a predetermined value, the connection is treated as impossible, and the connection to the voiceprint telephone call center 13A is terminated (step ST47).
[0030]
Next, processing on the voiceprint telephone call center 13A side when voice input is recognized from the caller side in the state connected to the voiceprint telephone call center 13A in this way will be described with reference to the flowchart shown in FIG.
[0031]
First, assuming that the voice input of “Mr. A” is “Mr. AA”, the voice signal of “Mr. AA” is input to the voice print telephone call center 13A connected in advance, and the voice data is received via the switching unit 40. Received by the unit 41 (see FIG. 2). In the audio data receiving unit 41, noise and the like are canceled through a filter or the like and sent to the voiceprint analyzing unit 42. The voiceprint analysis unit 42 analyzes the voiceprint and detects a voiceprint peculiar to “Mr. Ko” (steps ST50, ST51, ST52, ST53).
[0032]
In step ST53, when the detected voice data cannot be recognized as a specific person unique to “Mr. A”, the process goes to the caller side re-voice input processing routine. When it is recognized as a specific person, the specific person ("Mr. Ko") connection history database is searched (step ST54). In this connection history database, the name (full name) of the other party that the specific person ("Mr. Kou") has requested to connect to or has actually connected, or the pen name that is called in the normal conversation is stored. This is a configuration in which the order is sorted by the above method and a priority is assigned according to the frequency of use. This point will be described later with reference to FIG. For example, in the database shown in FIG. 6, {circle around (1)} a month b day c is connected to “Mr. C”, and the connection telephone number at that time is “aaa”. (2) At the time of a month b day d, connection is made with “Mr. D”, and the connection telephone number at that time is “bbbb”. (3) At the time of a month b day e, “AA” is connected, and the connection telephone number at that time is “cccc”.
[0033]
Therefore, “Mr. AA” included in the voice input data from the caller (“Mr. Ko”) this time is found to be the same person as “Mr. AA” at the time of e. It is possible to determine that there is a search target person, update the database, connect to the connection target person, and at the same time, call back the specified “Mr. A” information to the caller side (step ST55). , ST56, ST57).
[0034]
Here, if the search target person does not exist in the database, the process goes to the connection partner processing routine, and a procedure for connecting to the connection partner for the first time is performed (step ST58).
[0035]
When the connection with the other party is established, the switching unit 40 (see FIG. 3) is driven to switch the line and complete the connection with “Mr. AA” via the network. A conversation with “AA” is established (step ST59). In the case of absence, even if connection is possible, there is a so-called voice message recording function that accumulates voice data and transmits a message with a message and a message to the other party (step ST60). Further, in addition to the voice mail recording function, a voice reservation call function for calling at a predetermined time by analyzing voice data may be added. In the case of this voice reservation call, the caller side can be called back at a specific time before the reserved time. When connection is not possible, there are various requirements such as busy, line failure, absence of the other party, and connection mode processing is performed according to various conditions.
[0036]
Next, an accumulation method in the above-described specific person connection history database will be described with reference to FIG. This database has a configuration in which input voice data is decomposed and accumulated from the first voice print data to the nth voice print data for each wording, and is rearranged in descending order of frequency of use and prioritized. It has a structure in which ranking is given and responsiveness is made faster.
[0037]
For example, if the voice signal uttered by “Mr. Ko” is (1) “Mr. C”, “C” and “Mr.” are the first and second voiceprint data. {Circle around (4)} “O company company B” is the first voice print data, “B” is the second voice print data, and “san” is the third voice print data. In the search of the database stored in this way, for example, in the case of (4), if the first voiceprint data “Accounting” and the second voiceprint data “B” match, “Mr. B of accounting” is obtained. It is possible to call “eeee”. In this way, it is only necessary to decompose one sentence and the subject predicate and match the decomposed sentence. It should be noted that the database is not limited to such a search method, and may be any database corresponding to a request in words.
[0038]
As shown in FIG. 8, the caller side re-speech input processing routine is a routine for retrying the voice data of the specific person. First, the specific person's voice is adjusted (step ST65). This is to specify the specific person by changing the reference value of the voice data of the specific person acquired in advance, and the reference is based on the voice data of the specific person acquired in the past. Adjust the audio.
[0039]
When the voice cannot be specified even by adjusting the voice of the specific person, the voice is actually requested from the caller side, and in the embodiment, a vowel voice is requested (steps ST66 and ST67). Without being limited to this vowel, the specific sound data of the specific person can be adjusted by appropriately uttering a reference sound that matches the environment.
[0040]
When the vowel voice is input, the voice print based on the input vowel is compared with the voice print specifying data stored in advance (step ST69). If they are different from each other, the comparison with the voiceprint specifying data by repeatedly inputting vowels is performed until a predetermined count value is reached. Even if the voice print specifying data is repeatedly compared up to a predetermined count value, in the case of NG, it is determined that the specific person cannot be specified, and the process ends (steps ST70, ST71, ST72). In step ST69, if the input vowel voice data and the voiceprint identification data match, the slightly different voiceprint identification data that coincides is stored and used for later voiceprint identification (step ST70). ). In this way, by capturing and accumulating changes in the external environment and the voiceprint of the caller each time, the translation function described above can be strengthened, and voiceprint specific data having a wide range of voice changes can be obtained. Can do.
[0041]
Next, the connection partner side telephone number input routine generated in FIG. 6 will be described with reference to FIG.
[0042]
When the voiceprint analysis is completed and the other party's telephone number is not stored in the specific person connection history database, it is necessary to newly input voice and input the other party's telephone number. First, if the destination telephone number does not exist, the caller side is requested to input the destination telephone number (step ST75).
[0043]
In response to this request, when there is an input of the other party's telephone number from the sender side, the input telephone number is returned to check whether it is correct (steps ST76 and ST77).
[0044]
When the destination telephone number is correct, the process returns to FIG. 6 to update the voiceprint database and connect with the new telephone number (step ST78).
[0045]
In step ST78, if the other party's telephone number input by voice cannot be decoded or is an inappropriate telephone number, a response request is made again a predetermined number of times. (Steps ST79, ST80, ST81).
[0046]
In this way, the voiceprint telephone system is improved to a central translation system that recognizes the voiceprint telephone call center 13A (refer to FIG. 2) collectively, compared with the conventional system (terminal translation) that is individually recognized by the telephone. First, the voice print phone call center 13A is called, and the voice of the voice print phone call center 13A recognizes the voice and is immediately connected to Mr. A. The By doing so, there are the following merits.
[0047]
(1) Since the large-scale computer of the voice print telephone call center 13A is used, the storage capacity becomes infinite, and the recognition rate (translation ability) is greatly improved. (2) Even when the telephone 30 is lost or replaced with a new product, it is stored in the telephone directory (voice print database) at the voice print telephone call center 13A, so no change work is required. (3) The telephone book can be registered by voice by calling the voiceprint telephone call center 13A and anyone can use it easily. (4) The dial of the mobile phone becomes unnecessary, and it can be used even in places where it is unfamiliar with the operation of the dial or difficult to see. (4) A dial and buttons (tenkeys) are not required, and the telephone can be reduced in size, and a card type or wristwatch type telephone can be obtained.
[0048]
Next, a voiceprint telephone system according to a second embodiment will be described with reference to FIG.
[0049]
As shown in FIG. 10, the voiceprint telephone system according to the second embodiment is a system having an e-mail function in which voice data is converted into text, an e-mail document is created and transmitted to the other party, and the configuration is as follows. The voice print telephone 30, a network 31 comprising a switch network or the like that can be connected to the telephone, a voice print telephone call center 13 B that can be connected to the network 31, and receiving and displaying e-mails of an unspecified number of other parties. It is comprised from the information equipment 50 provided with the display which can do.
[0050]
The voiceprint telephone 30 has the configuration shown in FIG. 2B, and since it has already been described, its description is omitted.
[0051]
The voiceprint telephone call center 13B is a center that can respond to connection requests from an unspecified number of voiceprint telephones 30. The voiceprint telephone call center 13B is connected to the information device 50 on the other side requested by the voiceprint telephone 30, and the voiceprint telephone. Obtained by the voice data receiving unit 41 that receives the voice data from 30, the voice print analyzing unit 42 for analyzing the voice print of the received voice data and recognizing that it is specific voice data, and the voice print analyzing unit 42 Based on the specific voice data, the voice print search unit 43 for searching for a telephone number or the like connected to the information device 50 of the other party based on the voice data included therein, and a search by a request of the voice print search unit 43, Voiceprint database 44 in which data such as connection requests by specific persons having specific voice data is stored, and the keyword “E-mail” is recognized in the voice data. The voice data text conversion unit 51 that converts the voice data into text, and the partner connected to the information device 50 of the partner based on the data obtained by searching the specific person connection history database (see FIG. 7). The connection part 52 is comprised.
[0052]
The information device 50 is a device having at least a telephone function and having a display. For example, the information device 50 may have a configuration in which a telephone and a personal computer are combined, and the configuration is not determined.
[0053]
A general processing form in the voiceprint telephone system having such a configuration is that when the voiceprint telephone 30 on the calling side is turned on, the voiceprint telephone call center 13B is automatically called and when the voiceprint telephone call center 13B is connected, The voice data is received by the voiceprint telephone call center 13B, and voiceprint identification and decoding are performed.
[0054]
If the voiceprint decoding is successful, the contents included in the voice data, in this case, the keyword data “E-mail” are checked. If not, the process goes to another processing routine. If it is determined, the speech text included in the speech data is converted into a sentence and stored. At the same time, the specific person connection history database (see FIG. 7) is searched based on the data for specifying the other party, for example, the data “Mr. A”, and it is used in the past or inputted in advance. Search for "Mr. A"'s data and secure a phone number.
[0055]
The act of calling the phone of “Mr. A” on the other side starts, and the caller “Mr. A” connects to “Mr. A” on the called side by switching so that the calling side and the called side can be connected. Email data can be sent to the other party. Of course, the keyword “E-mail” can be changed as appropriate.
[0056]
The other voiceprint database, calling of the voiceprint telephone call center, and the like are the same as those of the voiceprint telephone system described in the first embodiment, and thus description thereof is omitted.
[0057]
Next, a voiceprint telephone system according to a third embodiment will be described with reference to FIG.
[0058]
As shown in FIG. 11, the voiceprint telephone system according to the third embodiment is a system having a word processor function capable of capturing and displaying text data obtained by converting voice data into text, and the configuration thereof is a voiceprint telephone. It comprises an information device 30A provided, a network 31 consisting of an exchange network or the like and capable of connecting to a telephone, and a voiceprint telephone call center 13C capable of connecting to the network 31.
[0059]
The information device 30A provided with the voice print phone is composed of a voice print phone and a device equipped with a display, for example, a personal computer, and the voice print phone has the structure described in FIG. 2B of the first embodiment. Since it has already been explained, its explanation is omitted.
[0060]
The voiceprint telephone call center 13C is a center that can respond to connection requests from an unspecified number of information devices 30A. The voiceprint telephone call center 13C receives a voice data receiving unit 41 that receives voice data from the information devices 30A, and a voiceprint of the received voice data. Based on the specific voice data obtained by the voiceprint analysis unit 42 and the voiceprint analysis unit 42 for analyzing and recognizing the specific voice data, the voice data included in the voiceprint analysis unit 42 Recognize the voiceprint search unit 43, the voiceprint database 44 that is searched by the request of the voiceprint search unit 43 and stores data such as connection requests by a specific person having specific audio data, and the keyword “word processor”. The document conversion unit 60 that converts the voice data into text to create text data and aligns the document, and displays the aligned document And a sound transmission unit 61. Show.
[0061]
A general processing form in the voiceprint telephone system having such a configuration is that when the information device 30A on the calling side calls the voiceprint telephone call center 13C and the voiceprint telephone call center 13C enters a connected state, the voice data is transmitted to the voiceprint telephone call center 13C. Received, voiceprint identification and decoding.
[0062]
If the voiceprint decoding is successful, the contents included in the voice data, in this case, the data of the keyword “word processor” are checked. If not, the process goes to another processing routine, and if it is determined that the word processor is “word processor”, the speech text included in the speech data is converted into text (created as text data) and stored in an aligned manner.
[0063]
In this way, the documented voice data is sent from the voice transmission unit 61 to the display of the information device 30A and displayed. Therefore, the device using the information device 30A utters the word “word processor” and then utters the word to be documented, and the word becomes text data and is displayed on the display. . Such a simple operation can be realized because the voice print telephone call center 13C professionally decodes the voice and converts the voice data into text data. It should be noted that the keyword “word processor” is not limited to this, and may be a keyword accumulated by the learning function, and the setting can of course be changed as appropriate.
[0064]
The other voiceprint database (specific person connection history database), calling of the voiceprint telephone call center, and the like are the same as those of the voiceprint telephone system described in the first embodiment, and thus description thereof is omitted.
[0065]
Next, a voiceprint telephone system according to a fourth embodiment will be described with reference to FIG.
[0066]
As shown in FIG. 12, the voiceprint telephone system according to the fourth embodiment is a system having an internet access function capable of capturing voice data and accessing the internet by voice. It comprises an information device 30A provided, a network 31 comprising a switching network or the like and capable of connecting to a telephone, and a voiceprint telephone call center 13D capable of being connected to the network 31.
[0067]
The information device 30A provided with the voice print phone is composed of a voice print phone and a device provided with a display. The voice print phone has already been described with reference to FIG. Is omitted.
[0068]
The voiceprint telephone call center 13D is a center that can respond to connection requests from an unspecified number of voiceprint telephones. The voiceprint telephone call center 13D analyzes the voiceprint of the received voice data and the voice data receiving unit 41 that receives voice data from the voiceprint telephone. A voiceprint analysis unit 42 for recognizing that the voice data is specific, and for accessing the Internet using the voice data contained in the voiceprint analysis unit 42 based on the specific voice data obtained by the voiceprint analysis unit 42 The voiceprint search unit 43 for searching for a telephone number or the like connected to the information device of the other party, and data such as a connection request by a specific person who has specific voice data are searched by a request from the voiceprint search unit 43. Voice access database 44 and Internet access unit 6 for recognizing the keyword “Internet” and accessing the Internet Internet access voice input unit 66 for inputting voice for access when accessing the Internet, and Internet access data for selecting data obtained by accessing the Internet based on the voice data and storing it in the Internet database 68 Unit 67 and a data transmission unit 69 for sending desired access data to the display side of the information device 30A.
[0069]
A general processing form in the voiceprint telephone system having such a configuration is that the originating information device 30A calls the voiceprint telephone call center 13D automatically or manually using a voiceprint telephone, and the voiceprint telephone call center 13D is in a connected state. Then, the voice data from the voiceprint telephone is received by the voiceprint telephone call center 13D, and the voiceprint analysis unit 42 performs voiceprint identification and decoding.
[0070]
If the voiceprint decoding is successful, then check the content contained in the voice data, in this case the keyword data “Internet”, if not, go to another processing routine and recognize it as “Internet” Connect to the network to access the Internet. The data accessed to the Internet is displayed on the display of the information device 30A via the data transmission unit 69. When the user responds to the display on the display by voice, the voice data in response is transmitted to the voice print telephone call center 13D, and the access is made as a response to the Internet.
[0071]
In this way, since data obtained by accessing the Internet is sequentially sent from the data transmission unit 69 to the display and displayed, those using a voice print phone uttered the voice of the keyword “Internet”. Later, if you speak a word you enter while watching the Internet information displayed on the display, you can access the Internet based on that word. Such a simple operation can be realized because the voice decoding and call data analysis function and the learning function are effective in the voiceprint telephone call center 13D. Of course, the keyword “Internet” can be changed to a favorite keyword.
[0072]
The other voiceprint database (specific person connection history database), calling of the voiceprint telephone call center, and the like are the same as those of the voiceprint telephone system described in the first embodiment, and thus description thereof is omitted.
[0073]
Next, a voiceprint telephone system according to a fifth embodiment will be described with reference to FIG.
[0074]
As shown in FIG. 13, the voiceprint telephone system of the fifth embodiment is a system capable of performing electronic commerce by voice by authenticating a specific person having a specific voiceprint, and the configuration includes a voiceprint telephone. Information network 30A, an exchange network or the like, and a network 31 that can be connected to the telephone, and a voiceprint telephone call center 13E that can be connected to the network 31.
[0075]
The information device 30A provided with the voice print phone is composed of a voice print phone and a device provided with a display, for example, a personal computer. The voice print phone has already been described with reference to FIG. 2B of the first embodiment. Therefore, the explanation is omitted.
[0076]
The voiceprint telephone call center 13E is a center that can respond to connection requests from an unspecified number of voiceprint telephones. The voiceprint telephone call center 13E receives voice data from the voiceprint telephone of the information device 30A, and the received voice data. A voiceprint analysis unit 42 for analyzing voiceprints and recognizing specific voice data, and a specific person is specified based on the specific voice data obtained by the voiceprint analysis unit 42 and stored in the voiceprint database 44. A voiceprint search unit 43 that searches and authenticates a recitation number, shopping information, past payment state information, and the like, a shopping access unit 70 that recognizes a commercial transaction keyword, for example, the keyword “shopping”, and accesses the virtual shopping 75, In order to access the virtual shopping 75 and visually check the state of the actual product A product search unit 71, a data transmission unit 72 that transmits product data obtained by accessing this virtual shopping to the information device 30A, and a settlement unit 73 that performs settlement when a desired product is searched and hit. It consists of and.
[0077]
A general processing form in the voiceprint telephone system having such a configuration is that when the voiceprint telephone of the information device 30A on the calling side calls the voiceprint telephone call center 13E automatically or manually, and the voiceprint telephone call center 13E enters a connected state, The voice data is received by the voice print telephone call center 13E, and the voice print analysis unit 42 performs voice print identification and decoding.
[0078]
If the voiceprint decoding is successful, the contents included in the voice data, in this case, the keyword data “shopping” of the commercial transaction keyword is confirmed. If not, the process goes to another processing routine, and “shopping”. If it is determined, after recognizing the shopping possible condition based on the shopping information and the settlement information among the contents searched by the voiceprint search unit 43, shopping If possible, connect to the network 31 to access the virtual shopping 75. Data accessed to the virtual shopping 75 is displayed on the display of the information device 30 </ b> A via the data transmission unit 72. When the user responds by voice to the display on the information device 30A, the voice data in response is transmitted to the voice print telephone call center 13E, which is a response to the contents displayed by accessing the virtual shopping 75.
[0079]
Thus, since the data obtained by accessing the virtual shopping 75 is sent from the data transmission unit 72 to the display of the information device 30A and displayed, the one using the voiceprint telephone is called “shopping”. If the word to be input is uttered while viewing the virtual shopping information displayed on the display after the keyword is uttered, the virtual shopping 75 can be accessed based on the word. Such a simple operation can be realized because a voice decoding and voice data analyzing function and a learning function work effectively in a voice print telephone call center. Of course, the keyword “shopping” of the commercial transaction keyword can be changed. For example, “I want YXXX (for men) YYY (sweater)”, “I want to buy YYY (sweater) for XXX (for men)”, “I want YYY (sweater)”, “I want to buy YYY (sweater)” If you want to shop in words like "XXXX" and "YYY", you can access Virtual Shopping 75 using the search input data and display the result on the display to select your favorite product. Can do. Of course, it is limited to shopping, but it can be applied to the entire business transaction.
[0080]
In this way, when the desired shopping is established and the product is hit, the settlement unit 73 needs to settle the product and the amount. This settlement is performed by means such as a recitation number or affiliated loan set in advance in the voiceprint telephone call center 13E. Accordingly, since the voiceprint telephone call center 13E is configured to always grasp the shopping information and settlement information related to the specific person specified by the voiceprint, even in the settlement, smooth settlement is performed using the information. Is called. When payment is completed, a series of commercial transactions is completed.
[0081]
Since the other voiceprint database (specific person connection history database), calling of the voiceprint telephone call center 13E, and the like are the same as those of the voiceprint telephone system described in the first embodiment, the description thereof is omitted.
[0082]
Next, a voiceprint telephone system according to a sixth embodiment will be described with reference to FIG.
[0083]
The voiceprint telephone system according to the sixth embodiment is a system for making a call by connecting to the other party by a voiceprint using a telephone set that is generally spread, and the configuration is at least as shown in FIG. Connecting to a telephone 31 (a mobile phone in the embodiment) 80 having a numeric keypad such as a microphone, a speaker, and a numeric keypad, a network 31 that can be connected so as to make a call to the other party such as an exchange network, and the like. A voice print telephone call center 13F that can be made, and a telephone set 80 (mobile phone in the embodiment) 80 of an unspecified number of other parties.
[0084]
The voiceprint telephone call center 13F is a center that can respond to connection requests from an unspecified number of telephones 80. From the telephone 80 and the switching unit 40 for connecting to the telephone 80a on the other side requested from the specific telephone 80. A numeric keypad connection unit 81 that responds to a numeric keypad, a voice data input unit 41 that receives voice data when connected to a specific telephone 80, and a specific voice data by analyzing a voice print of the received voice data. The voiceprint analysis unit 42 for recognizing that the phone number is recognized, and on the basis of the specific voice data obtained by the voiceprint analysis unit 42, the telephone number connected to the other party is searched based on the voice data included therein. The voice print search unit 43 and the search for the voice print search unit 43 search, and data such as a connection request by a specific person having specific voice data is accumulated. That a voiceprint database (specific party connection history database) 44, and a destination connection portion 45 that connects to the destination phone 80a based on the retrieved data.
[0085]
As shown in FIG. 15, a general processing form in the voiceprint telephone system having such a configuration is a numeric keypad for calling the voiceprint telephone call center from the telephone 80 owned by “Mr. Ko” on the calling side (step ST90). ).
[0086]
Then, when a calling operation is performed using the numeric keypad for calling the voiceprint telephone call center 13F via the exchange network, the dial response connection unit 81 is activated to establish a connection state between the telephone 80 and the voiceprint telephone call center 13F (step ST91).
[0087]
When the telephone 80 owned by “Mr. Ko” is connected to the voice print telephone call center 13F and voice for calling the other party is generated from the telephone 80, the voice data is received by the voice print telephone call center 13F, and the voice print analysis unit Voiceprint identification and decoding are performed by 42 (steps ST92, ST93, ST94).
[0088]
If the voiceprint decoding is successful, the voiceprint search unit 43 then searches the database based on data for specifying the other party included in the voice data, for example, data “Mr. A” (becomes a key term). The data of “Mr. A” used in the past or inputted in advance is searched. When a hit is found, the other party's connection unit 45 calls the other party's "Mr. A" phone (step ST95).
[0089]
The act of calling the telephone of “Mr. A” on the other side starts, and the switching unit 40 switches the connection between the calling side and the called side, whereby “Mr. A” on the calling side becomes “Mr. A” on the called side. The user can connect and have a conversation (step ST96).
[0090]
As described above, in the telephone 80 on the caller side, as long as the ten-key is entered and the voice print telephone call center 13F is connected, the call center automatically connects to the other party based on the voice request to the telephone. It is. Since the voiceprint database (specific person connection history database), the routine for inputting the other party's telephone number into the voiceprint database, and the like used for this connection are the same as those described in the first embodiment, the description thereof is omitted.
[0091]
In this way, by setting up the voice print telephone call center 13F, it is possible to use the telephones (black telephones, etc.) and mobile telephones that are widespread in ordinary homes and businesses to call the other party by voice without using a telephone number. You can call and have a conversation.
[0092]
In addition, if the voice print telephone described in the second to fifth embodiments is replaced with a general ten-key telephone, only the difference between calling the voice print telephone call center 13F by the ten key-in or by voice. Yes, after the connection is established, various functions such as voice e-mail (second embodiment), voice word processor (third embodiment), voice internet (fourth embodiment), voice The present invention can be applied to all electronic commerce (fifth embodiment).
[0093]
Next, a voiceprint telephone system according to a seventh embodiment will be described with reference to FIG.
[0094]
In the voiceprint telephone system according to the seventh embodiment, the analog voice signal that matches the voiceprint is recorded in advance, and the voice signal input from the telephone is compared with the matching analog voice signal. This is what we did.
[0095]
As shown in FIG. 16, the configuration for decoding a voiceprint using such an analog signal includes an analog signal generation unit 110 that inputs an audio from the telephone 100 and generates an analog signal, and the generated analog signal. An analog data storage unit 120 that stores signals, an A / D conversion unit 130 that converts the generated analog signal into a digital value, and an audio signal that is preliminarily considered to match the audio data converted into the digital value. A voiceprint matching circuit 140 to be compared, a comparison circuit 150 to compare a voice signal matched with a digital value with an analog signal, an analog voice matching circuit 160 to extract the voice signal when matched with an analog signal, and a matched analog voice signal Voice print matching analog data storage unit 170 for storing the analog voice signal and A / D conversion for converting the matched analog voice signal to digital And a section 180..
[0096]
In the decoding of a voice print by an analog signal having such a configuration, first, it is determined whether or not the input voice signal matches with a digital value. If they match, the voice print matching analog signal stored in advance as an analog signal is determined. Thus, the voice print matching operation can be performed with extremely high accuracy by comparing the voice prints by analog values and comparing the voice prints at a higher level.
[0097]
【The invention's effect】
As described above, the voiceprint telephone system according to the present invention provides a voiceprint telephone call center that recognizes a voiceprint of voice data and makes a call to the other party, so that a normal call can be made without selecting a telephone number or the like. There is an effect that it is possible to connect to a desired other party by voice data consisting of the words that appear.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram showing the overall configuration of a voiceprint telephone system according to the present invention.
FIG. 2 is an overall configuration diagram for telephone connection to the other party in the voiceprint telephone system according to the first embodiment;
FIG. 3 is a block diagram showing a general connection form for connecting to the other party via the voice print telephone call center.
FIG. 4 is a flowchart showing a calling routine from a voice print telephone to a voice print telephone call center.
FIG. 5 is a flowchart showing a calling routine from a voice print telephone to a voice print telephone call center.
FIG. 6 is a flowchart showing a processing routine in the voice print telephone call center.
FIG. 7 shows a part of the contents of the voice print database.
FIG. 8 is a flowchart showing the same caller side re-voice input processing routine;
FIG. 9 is a flowchart of the other party telephone number input routine.
FIG. 10 is a flowchart showing a processing routine of a voiceprint type voice email system in the voiceprint telephone system of the second embodiment.
FIG. 11 is a flowchart showing a processing routine of a voiceprint voice word processor system in the voiceprint telephone system of the third embodiment.
FIG. 12 is a flowchart showing a processing routine of a voiceprint voice internet system in the voiceprint telephone system of the fourth embodiment.
FIG. 13 is a flowchart showing a processing routine of a voiceprint type voice electronic commerce system in the voiceprint telephone system of the fifth embodiment.
FIG. 14 is an overall configuration diagram of a numeric keypad voice print phone system in a voice print phone system according to a sixth embodiment;
FIG. 15 is a block diagram showing a general connection form using the numeric keypad.
FIG. 16 is a principal block diagram for decoding a voiceprint using an analog signal in the voiceprint telephone system according to the seventh embodiment;
[Explanation of symbols]
10; Voiceprint phone, 11; Tenkey phone, 13; Voiceprint phone call center, 13A; Voiceprint phone call center, 13B; Voiceprint phone call center, 13C; Voiceprint phone call center, 13D; Voiceprint phone call center, 13E; Voice data, 15; voiceprint decoding, 16; translation function, 17; destination connection, 18; voice email, 19; voice word processor, 20; voice Internet, 21; voice electronic commerce, 30, voiceprint phone, 31; 32; Telephone, 33; Microphone, 34; Speaker, 35; Power means, 36; Connection section, 37; Data transmission / reception section, 40; Switching section, 41; Voice data reception section, 42; Voiceprint analysis section, 43; Part 44; voiceprint database (specific person connection history database) 45; Hand connection unit, 50; information device, 51; voice data sentence conversion unit, 52; partner connection unit, 65; internet access unit, 66; voice input unit for internet access, 67; internet access data unit, 68; internet database 69; data transmission unit, 70; shopping access unit, 71; merchandise search unit, 72; data transmission unit, 73; settlement unit, 80; telephone, 80a; telephone, 81; dial response connection unit, 100; Analog signal generation unit, 120; analog data storage unit, 130; A / D conversion unit, 140; digital voiceprint matching circuit, 150; comparison circuit, 160; analog voice matching circuit, 170; voiceprint matching analog data storage unit, 180 ; A / D converter

Claims (1)

  A telephone having at least a microphone, a speaker, and power means, and the telephone calls and connects a voice print telephone call center having a function of decoding a predetermined voice print when the power means can make a call. When voice data input from the microphone is decoded at the voice print telephone call center side and the decoded voice print is recognized as specific voice data, the voice data is stored in a predetermined database. A voiceprint telephone having a function of calling a destination based on a stored destination number corresponding to the stored voice data if it is stored as a result of the search system.

Images