JPH1168920A - Speech telephone number setting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To precisely set a telephone number though a user does not have the memory of the correct name of an opposite party at the time of recognizing speech at a speech telephone number setting device. SOLUTION: The speech telephone number setting device setting a telephone number to telephone by the speech of the name of an opposite side is provided with a microphone 1 for inputting speech, a single dictionary 3 storing telephone numbers by each name of opposite sides, dividing the respective names of the opposite sides into plural groups and registering speech patterns by each group of the divided order, a speech recognition processing part 2 successively collating speech inputted from the microphone with the speech pattern and recognizing it in the order of the group of word dictionaries, and a telephone number setting part 5 which sets a corresponding telephone number by judging the telephone number being specified when only one recognition is executed within a group by the speech recognizing processing part and monitoring the successive collation of a next group by judging the impossibility of the specifying of the telephone number when plural recognition are executed within the group on the other hand.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice telephone number setting device for recognizing a name of a partner by voice recognition and setting a telephone number. The present invention relates to a device that allows a telephone number to be set.

[0002]

2. Description of the Related Art In the above-mentioned voice telephone number setting device, a telephone number is set by voice input instead of a telephone number being set manually by dialing, a push button or the like. In the voice recognition word dictionary provided in the voice telephone number setting device, many voice patterns such as the first and last names of the other party are registered. When the voice input from the microphone and the voice pattern in the word dictionary are compared and matched, the first and last names of the other party are specified, and a telephone number is set for the specified first and last names of the other party. When the number of telephone numbers to be set increases, the family name may be commonly referred to (the same family name), and it is necessary to accurately identify the other party by the family name. When the number of telephone numbers to be set increases, the name of the first and last names may be common (the same name and the same name). Therefore, it is necessary to add the address to the first and last names to accurately identify the other party.

[0003]

However, when the number of telephone numbers increases in this way, the voice telephone number setting device described above, when the user calls the other party, if a part of the first and last name and address is incorrectly input, voice input is performed. On the contrary, there is a problem that the voice recognition is not correctly performed, the partner cannot be specified, and the telephone number cannot be set accurately. Especially,
Mistakes increase as the name and address are called later. For this reason, if the voice of the other party is input while looking at the telephone directory, the effect of setting the telephone number by voice is reduced by half.

Accordingly, the present invention has been made in view of the above problems, and provides a voice telephone number setting device capable of setting a telephone number by accurately inputting the first and last name and address of a partner even if the telephone number increases. The purpose is to:

[0005]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention provides a voice telephone number setting device for setting a telephone number to a telephone by voice of a name of a partner, a microphone for inputting voice, A word dictionary that stores a telephone number for each of the names, and divides each of the names of the other party into a plurality of groups, and registers a voice pattern for each of the groups in the divided order; A voice recognition processing unit for sequentially collating the voice input from the microphone with the voice pattern and recognizing that the telephone number is specified when only one recognition is performed in the group by the voice recognition processing unit; A corresponding telephone number is set. On the other hand, when a plurality of recognitions are performed in the group, it is determined that the telephone number cannot be specified, and the successive verification of the next group is monitored. And a telephone number setting unit.

Specifically, the word dictionary registers voice patterns in which the name of the other party is divided into two groups, a last name and a first name, and the telephone number setting unit recognizes a plurality of voice patterns in the last name voice pattern group. Is performed, the telephone number is specified by a group of first name voice patterns; voice patterns in which the other party's name is divided into three groups, a last name, a first name, and an address, are registered; When a plurality of recognitions are performed in the group of the name voice patterns, the telephone number is specified by the group of the address voice patterns.

By this means, if the same name increases due to an increase in the telephone number, the telephone number is specified by the first name, and if the same name increases, the telephone number is specified by the address. A telephone number can be set by voice input of name, address, and the like.

[0008]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram illustrating a voice telephone number setting device according to the present invention. As shown in the figure, when a voice is input to the microphone 1 in the voice telephone number setting device, the voice recognition processing unit 2 analyzes and encodes the characteristics of the voice input from the microphone 1. Word dictionary 3
In, the voice pattern encoded by the voice recognition processing unit 2 is registered as a database by dividing the name of the other party into a plurality of groups. Further, the voice recognition processing unit 2 compares the voice input from the microphone 1 with the voice pattern registered in the word dictionary 3 and checks which registered voice pattern matches. Phone number setting part 5
Has a voice synthesizing device (not shown), performs voice guidance with the speaker 7, and sequentially divides the other party's name into a plurality of groups based on the processing result of the voice recognition processing unit 2. A collation process is performed, the name of the other party is specified, and a telephone number is set to the telephone 6. The display unit 8 displays a database of the word dictionary 3 for confirmation. When the voice telephone number setting device is mounted on a vehicle, the display unit 8 is controlled based on a vehicle speed signal from a vehicle speed sensor 9.

The database of voice patterns registered in the word dictionary 3 has a structure. Specifically, as shown in FIG. 1, a memory name corresponding to one telephone number is 2
It has a double structure consisting of two voice patterns. For example, two memory names are a voice pattern A (Tanaka) of the same surname of group 1 representing the surname of the other party and voice patterns a1 (Ichiro) and a2 (Hanako) of the second group representing the name of each opponent.
These two memory names are associated with different telephone numbers.

The reason why the voice pattern B of the group 1 and the voice pattern b of the group 2 are double-structured is to consider that the same name will be obtained by a future addition. FIG. 2 is a flowchart illustrating the operation of the telephone number setting unit 5. As shown in the figure, the telephone number setting unit 5 accesses the group 1 of the word dictionary 3 in step S1.

In step S2, it is confirmed that the voice from the microphone 1 is processed by the voice recognition processing unit 2. In step S3, it is determined whether the recognition result is correct. If the recognition result is incorrect in step S4, the speaker 7 guides the user back as in "Please say your last name again" and returns to step S1. This is because voice input is performed again for erroneous recognition not related to the telephone number setting. If no reflection is required, the telephone setting process ends.

In step S5, if the recognition result is correct, it is determined whether the telephone number can be determined. That is, when the recognition result is the voice pattern B of group 1 (see FIG. 1), the telephone number can be determined because there is only one telephone number, but the recognition result is the voice pattern A of group 1. In this case, the telephone number cannot be determined because there are two telephone numbers.

If the telephone number can be determined in step S5, the telephone number is set to telephone 6 (step 6), and the process is terminated. If the telephone number cannot be determined in step S5, access is made to group 2 of the word dictionary (step 7). If there is no utterance within a predetermined time in step S8, the following inquiry is made, such as "Please utter your name."

In this inquiry, an object to be set can be directly inquired, "Please say a2 (Hanako)." In step S9, it is confirmed that the utterance obtained from the microphone 1 is processed by the voice recognition processing unit 2 in synchronization with the inquiry. In step S10, it is determined whether the recognition result is correct.

If the result of the recognition is incorrect in step S11, a return is made as in "Please say your name again" or "Please say a1 or a2" and return to step S7. If no reflection is required, the telephone setting process ends. If the recognition result is correct in step S12, it is determined whether the telephone number can be determined. If the telephone number can be determined, the process proceeds to step S6. If the telephone number cannot be determined, the process ends, and the process returns to step S1 and starts over again.

As described above, when a telephone number is set by voice input, voice recognition is performed in a double structure such as last name and first name. By synchronizing the next utterance, the telephone number can be determined accurately even if the user does not memorize the first and last names of the other party accurately. In addition, the user has been nervous when inputting voice to avoid making mistakes in the first and last names of the other party, as in the past. Was.

In the above, the case of the double structure of the first and last name of the individual has been described.
XX. A double structure such as a laboratory may be used. FIG.
FIG. 3 is a diagram illustrating another example of a database registered in the word dictionary 3 of FIG. 1. As shown in this figure, the memory name is
The voice pattern of group 3 relating to the address is added, and a triple structure of the last name, first name and address is made. This is because when the registration of the telephone number increases and the first and last names match, the telephone number of the other party is specified by the address (for example, the address in Tokyo or the address in Osaka).

In general, in response to a further increase in the registration of telephone numbers, it is possible to make the memory names more multiplexed according to the characteristics of the other party, such as hobbies and ages. In FIG. 3, an address inquiry step may be added. FIG. 4 is a view for explaining another example of the database registered in the word dictionary 3 of FIG. If the same person has two telephone numbers, as shown in the figure, the telephone numbers are compared with the voice pattern A of group 1 and the voice pattern a1 of group 2 having the same name. At the same time, time periods, for example, morning and afternoon are compared. In the word dictionary 3, time zones such as morning and afternoon are set by the ten keys 4. In addition, the telephone number setting unit 5
The time of the call is detected by a built-in timer, and a call is made to the place where the other party exists depending on the time zone. As a result, the operation of making a call can be reduced.

FIG. 5 is a view for explaining another example of the database registered in the word dictionary 3 of FIG. As shown in the figure, the word dictionary 3 is provided with a column for recording the frequency of making calls, in correspondence with the memory name of the same surname. Each time the telephone number is determined in steps S12 and S6 in FIG. 2 and the telephone number is set, the memory frequency of the other party specified in the column of the usage frequency for the first and last names (A, a1) and (A, a2) is displayed. The frequency is updated and recorded as N1, N2. For example, if the usage frequency of the first name and the last name for the voice pattern A of the group 1 is N1> N2, an inquiry to the user such as "Please say a2 (Hanako)" is made in the inquiry of step S8. If there is no utterance within a predetermined time, an inquiry such as "Please utter a1 (Ichiro)" is made.

As described above, the order of inquiries is learned in the order of frequency of use, and in the case of the same surname, the first inquiry involves the utterance of the name of the frequently used partner, thereby reducing unnecessary inquiries. . That is, the time to search can be reduced by weighting the database, and as a result, the time for telephone setting can be reduced. FIG. 6 is a flowchart for explaining a modification of FIG. 2 when there are many memory names with the same surname, and FIG. 7 is a diagram showing an example of a database registered in the word dictionary 3 when there are many memory names with the same surname. FIG. 8 shows an example of a list of memory names displayed on the display unit 8. FIG. 6 differs from FIG. 2 in step S7A.
Has been added. In step S7A,
As shown in FIG. 7, among the memory names registered in the word dictionary 3, all the surnames and first names belonging to the last name recognized in step S3 are displayed on the display unit 8 together with the frequencies in the order of frequency as shown in FIG. You.

Therefore, when there are a large number of memory names with the same surname, and when the required partner is low in the frequency order in the inquiry in step S8, the user can look at the display on the display unit 8 and speak, so that the telephone number can be set. Can be shortened. Also, since the user speaks while watching the display, the first and last names of the other party can be accurately input by voice. FIG. 9 is a diagram showing a modification of FIG. If there is a memory name having the same name and the same name, the word dictionary 3 is added with the voice pattern of the group 3 relating to the address in the database of FIG. If the telephone number cannot be determined even if the utterance is recognized in the access of the groups 1 and 2, all the addresses belonging to the first and last names recognized in the access of the groups 1 and 2 are displayed on the display unit 8.
Then, as shown in the figure, they are displayed together with the frequency in order of frequency.

FIG. 10 is a diagram showing a modification of FIGS. When the voice telephone number setting device is used in a vehicle, when the vehicle is running based on the vehicle speed of the vehicle speed sensor 9,
As shown in the figure, the amount displayed on the display unit 8 is reduced.
In this way, the time required for confirming the list display content during traveling is reduced, and the maximum time for moving the line of sight during driving is reduced.

[0023]

As described above, according to the present invention, if the same name increases due to an increase in telephone numbers, the telephone number is specified by first name, and if the same name increases, the telephone number is specified by address. Therefore, the telephone number can be set by accurately inputting the other party's last name, first name, address, and the like by voice.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a voice telephone number setting device according to the present invention.

FIG. 2 is a flowchart illustrating an operation of a telephone number setting unit 5;

FIG. 3 is a diagram illustrating another example of a database registered in the word dictionary 3 of FIG. 1;

FIG. 4 is a diagram illustrating another example of a database registered in the word dictionary 3 of FIG.

FIG. 5 is a diagram illustrating another example of a database registered in the word dictionary 3 of FIG. 1;

FIG. 6 is a flowchart illustrating a modification of FIG. 2 when there are many memory names with the same surname.

FIG. 7 is a word dictionary 3 when there are many memory names of the same surname
FIG. 4 is a diagram showing an example of a database registered in the database.

FIG. 8 is an example of a list of memory names displayed on a display unit 8;

FIG. 9 is a diagram showing a modification of FIG. 8;

FIG. 10 is a diagram showing a modified example of FIGS. 8 and 9;

[Explanation of symbols]

 Reference Signs List 1 microphone 2 voice recognition processing unit 3 word dictionary 4 numeric keypad 5 telephone number setting unit 6 telephone 7 speaker 8 display unit 9 vehicle speed sensor

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification code FI G10L 3/00 561 G10L 3/00 561G

Claims (10)

[Claims] 1. A voice telephone number setting device for setting a telephone number to a telephone by voice of a name of a partner, a microphone for inputting voice, a telephone number for each name of the partner, and each of the names of the partner Into a plurality of groups, and a voice dictionary for registering voice patterns for each of the groups in the order in which the voices were voiced, and voices that are sequentially collated with the voice patterns to recognize voices input from the microphone in the order of the word dictionary groups. A recognition processing unit, when only one recognition is performed in the group by the voice recognition processing unit, it is determined that the telephone number has been specified, and a corresponding telephone number is set; A telephone number setting unit that determines that the telephone number cannot be specified when the call is made, and monitors the successive verification of the next group. No. setting device. 2. The word dictionary registers voice patterns obtained by dividing the other party's name into two groups, a last name and a first name, and the telephone number setting unit performs a plurality of recognitions on the last name voice pattern group. 2. The voice telephone number setting device according to claim 1, wherein in the case, the telephone number is specified by a group of voice patterns of names. 3. The word dictionary registers voice patterns in which the other party's name is divided into three groups, a last name, a first name, and an address. When the recognition is performed, the phone number is identified by a group of voice patterns of the address,
The voice telephone number setting device according to claim 1. 4. The telephone number setting unit further includes a voice synthesizing device, wherein the voice synthesizing device guides the sequential collation by the voice recognition processing unit by voice.
The voice telephone number setting device according to claim 1. 5. When the word dictionary stores a telephone number for a voice pattern of the same partner's name in association with a time zone, the telephone number setting unit sets the telephone number based on the time at which the telephone number is set to a telephone. The voice telephone number setting device according to claim 1, wherein the telephone number is specified by specifying a time zone. 6. The voice recognition process according to claim 1, wherein the word dictionary is provided with a use frequency field for storing a frequency at which a telephone number is set by the telephone number setting unit for a name of a partner having a common voice pattern in the group. The unit performs input voice matching based on the frequency of the voice pattern,
The voice telephone number setting device according to claim 1. 7. The word dictionary is provided with a use frequency field for storing a frequency at which a telephone number is set by the telephone number setting unit for a name of a partner having a common voice pattern in a group, and The voice telephone number setting device according to claim 1, wherein the unit has a display unit, and displays the name of the other party based on the frequency of the voice pattern. 8. The word dictionary stores, when a voice pattern in which the name of the other party is divided into two groups, a last name and a first name, is registered, each of the names of the other party with respect to the last name having a common voice pattern is registered. The voice telephone number setting device according to claim 7, wherein a frequency is assigned. 9. The word dictionary, when a voice pattern in which the name of the other party is divided into three groups of a last name, a first name, and an address is registered, stores the name of the other party with respect to the first and last names having a common voice pattern. The voice telephone number setting device according to claim 7, wherein a frequency is assigned to each of them. 10. The voice telephone number according to claim 7, wherein when the voice telephone number setting device is mounted on a vehicle, the display amount of the display unit is reduced when the vehicle is running. Setting device.