JPH02117252A - Telephone set - Google Patents

Abstract

PURPOSE:To prevent erroneous dialing from being performed and to improve operability by issuing a call to a recognition candidate with the highest order of candidate displayed in voice input, stopping the issuing of the call when the selection of another recognition candidate is performed while issuing the call, and issuing the call by a selected recognition candidate. CONSTITUTION:An operator raises a handset 2 and performs the input of an opposite name in voice according to the message of a display part 7, and it is supplied to a vocalizing recognition device 10, and a prescribed number of candidates of the opposite name is found. A CPU 1 retrieves a storage part 6, and reads out the corresponding telephone number of an opponent and attached, information, and displays them on the display part 7, and also, outgoing to the opponent that is a first candidate out of the candidates of the opposite name is started. The opposite name and the attached information on the display part 7 are checked during a line is connected. and when they are wrong, desired opposite name and information out of the opposite names and the attached information can be selected by a selection button. The CPU 1 interrupts the work of line connection under execution while the above selection is being perform ed, and starts the outgoing to the opponent selected newly.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention (Field of Industrial Application) The present invention relates to a telephone capable of automatic dialing by voice input.

(Prior Art) In recent years, many electronic devices have become ubiquitous around us. Many of these devices can have multiple functions or operations. Although this is very convenient, it is often inconvenient in operation. From this perspective, it has become very important to consider the relationship between various electronic devices and the people who use them.

To explain this in detail, for example, in the case of a telephone, with the advent of automatic electronic exchanges, it has become possible to make a call to the other party by simply dialing the other party's phone number, without having to go through the trouble of having a switchboard handle the transfer operations. Ta. Such automatic functions are very convenient. However, such an automatic function can be used only when the telephone number of the other party is known, and cannot be used when the telephone number of the other party is not known. Of course, telephone directories and the like are prepared for such cases, but telephone directories contain a lot of information that is unnecessary for users and are generally difficult to use.

Therefore, the inventors of the present invention determined a predetermined number of recognition candidates from among a plurality of recognition candidates obtained by recognition processing of input speech indicating the other party, in descending order of candidate ranking, and selected the candidates according to a selection instruction from the user. We are proposing a telephone that automatically dials the called party.

However, with these automatic dialing telephones, the destination is selected from among multiple recognition candidates according to the user's selection instructions, so while it is possible to prevent incorrect dialing caused by voice recognition processing, it is difficult for the user to always Since it forces a selection operation, it is often troublesome for the user.

(Problem to be Solved by the Invention) As described above, the above-mentioned automatic dialing telephone can automatically dial a destination by inputting the destination name by voice, but incorrect dialing caused by voice recognition processing can occur. In order to prevent such problems, the user selects the other party from among multiple recognition candidates based on the user's selection instructions, which forces the user to constantly select, which can be troublesome for the user. There is.

The present invention was made in consideration of the above circumstances, and an object of the present invention is to provide a telephone that can improve operability while preventing erroneous dialing caused by automatic dialing using voice recognition processing. There is.

[Structure of the Invention] (Means for Solving the Problems) The present invention provides a voice input means into which voice is input, a voice input by the voice input means, and a plurality of recognition candidates corresponding to the voice. and a voice recognition means for outputting a candidate ranking corresponding to each of these recognition candidates, a display means for displaying two or more recognition candidates among the plurality of recognition candidates output from the voice recognition means, and the voice recognition means a first calling means that calls a recognition candidate with the highest candidate ranking among the recognition candidates displayed by the display means based on the ranking candidates output from the first calling means; selecting means for selecting a recognition candidate other than the recognition candidate to which the call was made based on the result displayed on the display means; and while the first calling means is performing the calling operation; and a second calling means that stops the calling operation of the first calling means and makes a call based on the recognition candidate selected by the selecting means when the selection means makes a selection. , which solves the above-mentioned problems.

(Function) That is, according to the present invention, when a voice is input, a plurality of recognition candidates and candidate rankings corresponding to these recognition candidates are output, a plurality of recognition candidates are displayed, and first, the candidate ranking is the highest. A call is made to a high recognition candidate. If another recognition candidate is selected while this calling operation is being performed, the calling operation is stopped and a call is made using the selected recognition candidate.

(Example) Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing a schematic functional block configuration of a telephone according to an embodiment of the present invention. This telephone mainly includes a control device 1 consisting of a CPU, and includes various parts described below. That is, this telephone is connected to a telephone line via a line interface 8, and displays information such as the sending/receiving item temperature 2, a button input section 5 including a dial pad or selection buttons lla and llb, and a telephone number with the name of the other party as a heading. an information storage unit 6 storing information, a liquid crystal display or a transmission indicator lamp 12 a s 12 b s 12 c
a display section 7 consisting of a speech recognition section 3 and a recognition dictionary 4;
The speech recognition device 10 is configured to include a speech recognition device 10 consisting of the following. Second
The figure shows the external configuration of a telephone equipped with such functions.

Now, the voice recognition device 10 recognizes the characteristics of the word voice indicating the name of the other party inputted via the handset 2 by the voice recognition unit 10.
The input speech word (name of the other party) is recognized by analyzing and extracting the word and comparing it with each of the phonetic characteristics (standard patterns) of a plurality of other parties registered in advance in the recognition dictionary 4.

The control device 1 obtains a predetermined number of recognition candidates (destination name candidates) for the input speech recognized by the speech recognition device 10, for example, from among the plurality of recognition candidates (destination name candidates) for the destination, and makes a selection as described later. Submitted for processing.

On the other hand, in the information storage unit 6, the telephone number of the other party to be recognized is registered in advance in correspondence with the other party to be recognized. Of course, the information storage unit 6 may store additional information regarding the other party, such as affiliated member's name, position, address, name, occupation, etc., if necessary.

The button input section 5 constitutes a dial pad [1, 2,
~9,0. *, #J buttons and selection buttons lla, l
Consists of lb etc. Each of these buttons sends a signal corresponding to the button to the control device 1 when pressed. This button input section 5 is also used for normal telephone operations (dial operation: making a call), but in this embodiment, it is also used for inputting information into the information storage section 6. That is, for example, 101 Japanese kana characters are assigned to one or more combinations of buttons as shown in FIG. 3, and information can be input by operating these buttons.

Note that the speech recognition device 10 may be provided with a function of recognizing monosyllables, and the information may be input in units of kana by voice. Of course, when inputting numerical information to the information storage section 6, the button input section 5 or the voice recognition device 10 may be used. Also, select button lla
, llb are provided corresponding to the second and third destination name candidates among the destination name candidates up to the third candidate displayed on the display unit 7 at the time of making a call.

As will be described later, when these selection buttons lla and 11b are pressed during a call to the first destination name candidate, the call is switched to the corresponding destination name candidate. ing.

On the other hand, the display section 7 includes a liquid crystal display, transmitting water lamps 12a, 12b, 12c, etc., and visualizes various display information and presents it to the user according to instructions from the control device 1. On the display section 7, a predetermined number of 10 candidate names among the recognition candidates recognized by the voice recognition device 10 and information such as telephone numbers corresponding to the parties are classified and displayed by item. Furthermore, various messages are also displayed on the display section 7, as will be described later. Furthermore, the calling lamps 12a, 12b, and 12C are provided corresponding to the destination name candidates up to the third candidate displayed on the display unit 7 when making a call. When a call is made, the calling lamps 12a, 12b, and 12c corresponding to the called party are turned on or blinking.

The line interface 8 exchanges signals between the telephone and the line under the control of the control device 1, and forms a communication path with the other party's telephone via the line.

Next, the operation of the telephone with the above configuration will be explained.

When attempting to make a call, first pick up the handset 2 as usual. Then, the display section 7 will display, for example, ``Please say the name of the other party.'' ” message is displayed. The operator follows this and inputs the name of the other party by voice. The voice input in this way is given to the voice recognition device 10 via the handset 2,
A predetermined number (for example, top three) of recognition candidates (destination name candidates) are obtained. The control device 1 searches the information storage section 6 for each of these destination name candidates, reads out the telephone number and attached information of each corresponding destination, and displays them on the display section 7, and also displays recognition candidates (destination name candidates). ) middle, 1st
Start calling the candidate destination. That is, a signal corresponding to the first candidate's destination telephone number is sent to the line interface 8.
to the line via. The line connection is then made in the same way as normal telephone operations. Although it takes some time until the line is actually connected, the operator confirms the name of the other party or the attached information displayed on the display section 7 during that time. At this time, if the destination is incorrect, the operator selects the desired one from the remaining recognition candidates (destination name candidates) or attached information displayed on the display unit 7 by pressing the button l.
Select using la and llb.

In response to this selection, the control device 1 cancels the line connection work in progress and starts calling the newly selected destination.

FIG. 4 is a diagram showing an outline of the above operation processing procedure.

This processing procedure begins with the input of voice indicating the name of the other party (the name of the destination party) from the above-mentioned handset 2 (step a). A voice signal input from the handset 2 is digitized, for example, by an A/D converter (not shown), and then supplied to the voice recognition section 10.

When the voice is input from the handset 2 in this way, the voice recognition process by the voice recognition section 10 is started (step b). This speech recognition process first extracts a speech section from an input signal, then calculates changes in the audio power and frequency analysis results of the input signal for that speech section, and extracts its acoustic features as an input speech pattern. It is done because of this. Then, the input speech pattern extracted through this acoustic analysis and the standard pattern of the destination name registered in advance in the recognition dictionary 4 are compared by determining the degree of similarity, etc., and recognition candidates for the input speech are obtained. . The recognition dictionary 4 involved in the recognition process of this input voice performs statistical processing on voice data collected from, for example, a large number of unspecified speakers, and calculates each of the names of a plurality of recipients to be recognized. This is a registered standard pattern.

However, in the case of a telephone that has a limited number of users, voice data is collected from a specific speaker (single or plural) using the telephone, and a standard pattern created from this collected data is registered in the recognition dictionary 4. You may leave it as is.

The matching process is performed by calculating the degree of similarity or distance between the input speech pattern and the standard pattern. Then, from the comparison results, those with high recognition candidate rankings, for example, the first to third ranking candidates are extracted as recognition candidates (destination name candidates) for the input voice described above (step C). In order to extract multiple destination name candidates, it is difficult to accurately recognize all the words in normal input speech from the viewpoint of recognition accuracy of word speech, but This is based on the fact that if you focus on the above, there is a high probability that the correct recognition result will be included.

After performing recognition processing on the input speech (word speech indicating the name of the other party) in this manner, the aforementioned control device 1 determines whether or not a recognition result has been extracted (step d).

If it is confirmed that the input voice has been rejected as a result of this determination, that is, if a recognition candidate cannot be extracted, the display unit 7 under the control of the control device 1
is used to display a message indicating that the voice recognition has been unsuccessful, for example, ``Please speak again.'' to prompt the user to re-enter the other party's name (step e). The voice recognition process is similarly performed on the voice signal re-inputted by this message.

In addition, if the recognition process for the repeatedly input voice ends in failure, for example, when the voice is repeatedly input a predetermined number of times, automatic dialing cannot be performed, and the telephone number list is checked and the button input section 5 It may be possible to output a message urging the user to dial by operating the , or to display a message requesting that another person take the place of the speaker.

Note that the message output here may be outputted as a sound instead of a display output by the display unit 7, or a message display output and an audio output may be used together. Regarding this voice output, for example, predetermined message data indicated by a character code string may be synthesized into voice using voice rule synthesis, etc., and then output.

Now, when destination name candidates are obtained as a plurality of recognition candidates for the input voice, the control device 1 searches the information storage unit 6 for these destination names (step f). This information storage section 6 stores attached information such as a telephone number and the name of the company to which the person belongs, with the name of the recipient as a heading. Control device 1
retrieves and extracts telephone numbers and attached information from such information storage unit 6 according to the above-mentioned destination name candidates. Then, the searched and extracted destination name and telephone number are displayed on the display section 7 attached thereto as necessary (step g).

Specifically, for example, when a user utters the input voice "Kato," the top three recognition candidates (destination name candidates) are determined by voice recognition: "Kato,""Sato," and "Saito." In this case, the numbers of the respective destination names are obtained from the information storage unit 6 as r03-365-5049J r044-655-3021J r045-321-6025J and displayed as shown in FIG.

Subsequently, the control device 1 performs automatic dialing by sending a signal corresponding to the telephone number of the first candidate to the line via the line interface 8 (step h). At this time, in order to indicate that automatic dialing (calling) is in progress, the corresponding calling indicator lamp 12a provided on the left side of the display section 7 is turned on or blinks. Incidentally, even if such a call-in-progress indicator lamp is not provided, it is also possible to indicate that a call is in progress by displaying the information attached to the destination name and dial in reverse or blinking.

The operator checks whether the calling party is correct by checking the blinking of the lamp indicating that the call is in progress (step i). If it is correct, wait for the line to be connected. After a while, the line will be connected and you can talk (step m)
becomes. On the other hand, if the I11 hand name is incorrect, the operator presses the selection button provided corresponding to the desired 111 hand name from among the remaining partner name candidates displayed on the display section 7. When a selection button is pressed, selection instruction information input processing is performed according to the button (
Step j). By inputting this selection instruction information, the automatic dialing in step h is first canceled (step k) L, then the telephone number of the selected destination is read out from the information storage section 6, and then the telephone number is read out from the information storage section 6, and then the telephone number The telephone number is sent and automatically dialed (step l).

As a result, this telephone and the other party's telephone are connected via an exchange (not shown) (step m), and the input/output audio of the handset 2 is exchanged with the other party's telephone from the line interface 8 via the exchange. It becomes possible to make a call.

Thus, according to the telephone set according to the present embodiment, the first candidate destination through voice recognition processing is automatically dialed immediately, eliminating the trouble of pressing the selection button one by one.

Additionally, even if the first choice is incorrect, you can check the second and third candidate names and phone numbers, and then select and instruct the desired recipient name and phone number to automatically dial. Therefore, it is possible to prevent incorrect dialing caused by automatic dialing using voice recognition processing.

Note that the present invention is not limited to the embodiments described above.

For example, when selecting the second candidate and the third candidate, as shown in FIG. 5, the corresponding selection buttons 1.1. a
, llb, but the selection is not limited to this. For example, assume that the results shown in FIG. 6 are obtained as recognition candidates up to the third candidate. At this time, by pressing the next candidate button, the destination of the call will be selected in order (1st candidate, 2nd candidate, 3rd candidate,
The first candidate, . . . ) may be selected, or the fourth candidate, the fifth candidate, . . . may be further displayed and selected.

Furthermore, the selection of destination name candidates is not limited to the use of the selection buttons described above. For example, instead of a selection button, each display item may be marked with ``■'', ``■'', and ``■'' for identification, and the selection may be made by pressing the corresponding key on the key pad of the button input section 5. “Ichi” again
``Second'' An input voice such as F San J may be recognized and the selection information may be input.

[Effects of the Invention] As explained above, according to the present invention, when a voice is input, a plurality of recognition candidates and candidate rankings corresponding to these recognition candidates are output, a plurality of recognition candidates are displayed, and first, , a call is made to the recognition candidate with the highest candidate rank. If another recognition candidate is selected while this calling operation is being performed, the calling operation is stopped and a call is made using the selected recognition candidate. Therefore, it is possible to improve operability while preventing erroneous dialing caused by automatic dialing using voice recognition processing.

[Brief explanation of the drawing]

FIG. 1 is a functional block diagram showing a schematic configuration of a device according to an embodiment of the present invention applied to a telephone, FIG. 2 is a diagram showing an example of the configuration of a front panel of a telephone according to an embodiment, and FIG. A diagram showing an example of assigning keys to kana for inputting information, FIG. 4 is a diagram showing the flow of processing procedures according to the embodiment, and FIGS. 5 and 6 respectively show speech-recognized destination name candidates and their destinations. FIG. 4 is a diagram illustrating an example of displaying a telephone number of a person, a display example of a telephone number of the other party, and an example of a form of a selection instruction. DESCRIPTION OF SYMBOLS 1... Control device, 2... Handset, 3... Voice recognition part, 4... Recognition dictionary, 5... Button input part, 6...
... Information storage unit, 7... Display unit, 8... Line interface, 10... Voice recognition device, lla, 1b.
...Selection button, 12 a s 12 b 112 c
...Transmission indicator lamp. Applicant Toshiba Corporation Patent Attorney Sasa Suyama - Figure 4 Button Mitsunobu Nakao'! ! - Nine telephone numbers 5th button dialing, Xianxian 7th number 6th figure

Claims (1)

[Claims] (1) A voice input means into which voice is input, and a voice recognition system that recognizes the voice input by this voice input means and outputs a plurality of recognition candidates corresponding to the voice and a candidate ranking corresponding to each of these recognition candidates. means, a display means for displaying two or more recognition candidates out of the plurality of recognition candidates output from the voice recognition means, and display means for displaying two or more recognition candidates out of the plurality of recognition candidates output from the voice recognition means, based on the ranking candidates output from the voice recognition means. a first calling means that calls the recognition candidate with the highest candidate rank among the recognition candidates that have been selected; and displaying the recognition candidates other than the recognition candidates to which the first calling means has called. selection means for making a selection based on the result displayed on the means; and when the selection by the selection means is made while the first calling means is performing a calling operation;
and a second calling means that stops the calling operation of the calling means and makes a call using the recognition candidate selected by the selection means.