JPH0950291A - Voice recognition device and navigation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve handleability in a voice recognition device. SOLUTION: This device is provided with a voice recognition processing means for estimating a vocabulary uttered by a user based on an inputted voice signal S5 and forming a candidate list in which the estimated vocabularies are arranged in order of probability and a control means 4 for displaying the candidate list on a display means 6 after executing a processing based on a first candidate among the candidate list and enabling the user to select the vocabularies on and after the second candidate, and executing the processing based on the vocabulary when the covabulary on and after the second candidates is selected. When the vocabulary of the first candidate is erroneously recognized, the erroneous recognition is instantly and easily corrected without executing the processing again from the beginning as conventional one, and the handleability, is improved as compared with conventional one.

Description

Detailed Description of the Invention

[0001]

[Table of Contents] The present invention will be described in the following order. Technical Field to which the Invention pertains Related Art Problems to be Solved by the Invention Means for Solving the Problems Embodiments of the Invention (FIGS. 1 to 5) Effects of the Invention

[0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice recognition device and a navigation device, and is suitable for application to, for example, an on-vehicle navigation device.

[0003]

2. Description of the Related Art Today, GPS (Global Positioning Sys
tem) Navigation devices that use radio waves from satellites to display the current position and direction of travel on a map have been put into practical use. In particular, an in-vehicle navigation device that displays the current position of an automobile is less troublesome than the conventional case where the user determines the current position by comparing the surrounding situation with a map book, and has been extremely difficult in recent years. Is popular in

This on-vehicle navigation system is equipped with a GPS
It receives radio waves from satellites and calculates the current position of the vehicle based on the received radio waves. And CD-ROM (Compact
Read the map data from DiskRead Only Memory),
The calculated position data is combined with the read map data and displayed on the display. As a result, the user can confirm the current position of the vehicle just by looking at the display, and can easily confirm the current position while driving.

[0005]

In the in-vehicle navigation device having such a structure, for example, when displaying a desired map, various operation switches provided on the remote commander or the main body must be operated. In this case, the user has to operate while watching both the remote commander (or the operation switch of the main body) and the display, which may make it difficult to concentrate on driving the vehicle.

Therefore, in recent years, a vehicle-mounted navigation device has been proposed which can be operated by a voice uttered by a user. This in-vehicle navigation device is provided with a voice recognition device, which recognizes the voice uttered by the user and operates based on the recognition result. However, in the in-vehicle navigation device using this voice recognition device, if erroneous recognition occurs, it is necessary to restart the operation from the beginning, or there is a disadvantage that it is necessary to utter again many times, There are still insufficient points in terms of usability.

The present invention has been made in consideration of the above points, and an object thereof is to propose a voice recognition apparatus and a navigation apparatus using the same, which can improve usability.

[0008]

In order to solve such a problem, in the present invention, a vocabulary uttered by a user is predicted based on an input voice signal, a plurality of predictable vocabularies are calculated, and a plurality of calculated vocabulary is calculated. After performing a process based on the vocabulary of the first candidate among the voice recognition processing means for creating and outputting a candidate list in which the vocabulary is arranged in a probable order, the candidate list is displayed. The vocabulary after the second candidate is displayed on a predetermined display unit so that the vocabulary after the second candidate can be selected by the user. When the vocabulary after the second candidate is selected by the user, the vocabulary based on the selected vocabulary is used. A control means for performing processing is provided.

After performing the processing based on the vocabulary of the first candidate, the candidate list is displayed on the display means so that the vocabulary of the second and subsequent candidates can be selected by the user, and the second candidate is selected by the user. If the subsequent vocabulary is selected, the processing based on that vocabulary is performed, so that if the vocabulary of the first candidate is erroneously recognized, there is no need to redo from the beginning as in the past, and immediately, And the misrecognition can be easily corrected.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

In FIG. 1, reference numeral 1 denotes an on-vehicle navigation device to which the present invention is applied as a whole, and an electric wave transmitted from a GPS satellite is received by an antenna 2 and a reception signal S1 thereof is sent to a GPS receiving section 3. It is designed to be entered. The GPS receiving unit 3 calculates the current position (specifically, latitude and longitude) based on the input reception signal S1 and outputs the calculated position data S2 to the arithmetic processing unit 4.

The arithmetic processing unit 4 firstly, based on the position data S2, the map data S3 around the current position from the CD-ROM 5
Is read. Next, the arithmetic processing unit 4 combines the read map data S3 with a mark indicating the current position to display data S4.
Is generated and the display data S4 is output to the display 6. The display 6 is, for example, a liquid crystal display, and displays an image based on the supplied display data S4. As a result, an image in which the current position (specifically represented by a mark) is superimposed on the map is displayed on the display 6.

The operation processing unit 4 also includes a remote commander 7
Are connected by infrared rays or wires, and various operation information and set values of the soft switch are input from the remote commander 7 to the arithmetic processing unit 4. Based on the input operation information, the arithmetic processing unit 4 enlarges the map or displays a map of the area desired by the user on the display 6, and software for determining the operation mode of the vehicle-mounted navigation device. Change the switch setting value.

By the way, this on-vehicle navigation device 1
Is provided with a voice recognition processing unit 8 so that operation information can be input also by voice.
The voice recognition processing unit 8 is provided with a microphone 9 for inputting a voice uttered by a user and a talk switch 10 for notifying the voice recognition processing unit 8 of the voice section.

When the user inputs the operation information by voice, the user first turns on the talk switch 10 and puts the voice recognition processing unit 8 into the voice detection state. The user then speaks into the microphone a vocabulary representing the operation information.
This voice is converted into an electric voice signal S5 by the microphone 9 and input to the voice recognition processing unit 8. The voice recognition processing unit 8 recognizes the vocabulary uttered by the user based on the input voice signal S5, and outputs the recognition result S6 to the arithmetic processing unit 4. In this case, the voice recognition processing unit 8 obtains the vocabulary that can be predicted as the most probable vocabulary as the first candidate, for example, up to the fifth candidate, and outputs the obtained candidate list as the recognition result S6. That is, the voice recognition processing unit 8 predicts a vocabulary spoken by the user, obtains a plurality of predictable vocabularies, arranges the obtained plurality of vocabularies in a certain order to create a candidate list, and outputs the candidate list as a recognition result S6. To do.

The arithmetic processing unit 4 discriminates the operation information based on the most probable vocabulary in the first candidate in the candidate list, and performs the processing based on the discriminated operation information. If the candidate list switch, which is a soft switch, is set to the ON state in advance, the calculation processing unit 4 displays the candidate list on the display 6 for a predetermined time, and the user operates the remote commander 7 during the display. If the vocabulary after the second candidate is selected, the processing based on the selected vocabulary is performed. If no selection operation is performed during the display, the process returns to the process based on the first candidate.

The reason why the candidate list of predictable vocabulary is displayed so that the user can select the vocabulary of the second and subsequent candidates is that the vocabulary of the first candidate is erroneously recognized. This is because there is a high possibility that the vocabulary originally uttered by the user is included in the vocabulary after the second candidate.
That is, in this in-vehicle navigation device 1, by displaying a candidate list of predictable vocabulary and selecting a vocabulary from the candidate list, even if misrecognition occurs, it can be immediately and easily performed. You can correct misrecognition,
It is possible to avoid redoing the operation from the beginning or re-speaking many times as in the conventional case.

The operation information input by voice described above will be described with reference to the flow chart shown in FIG.
First, in step SP2, the user turns on the talk switch 10 at step SP2. As a result, the voice recognition processing unit 8 enters the voice detection state. Next, in step SP3, the user utters a vocabulary representing the operation information. This voice is transmitted by the microphone 9 as a voice signal S5.
And is sent to the voice recognition processing unit 8. Next, in step SP4, the user turns off the talk switch 10 and puts the voice recognition processing unit 8 into the voice detection end state.

Next, in step SP5, the voice recognition processing section 8 voice-recognizes the vocabulary uttered by the user based on the voice signal S5. At that time, the speech recognition processing unit 8 obtains the vocabulary up to the fifth candidate with the most probable vocabulary as the first candidate, and the obtained candidate list is used as the recognition result S6.
To the arithmetic processing unit 4.

Next, in step SP6, the arithmetic processing section 4 determines operation information (for example, displaying a map of a predetermined area or enlarging the map) based on the vocabulary in the first candidate in the candidate list. Then, in the next step SP7, processing based on the determined operation information is performed. that time,
The arithmetic processing unit 4 displays the candidate list on the display 6 for a predetermined time if the candidate list switch is in the ON state, and if there is no selection operation by the user during that time, the processing proceeds to step SP8 to end the processing, and the selection operation is performed. If there is, the processing based on the selected vocabulary is performed, and the process proceeds to step SP8 to end the processing. Then, when the operation information is input again by voice, the operation is started from step SP1 and the processing is repeated.

Here, for example, the processing procedure of the arithmetic processing unit 4 (the portion of step SP7 in FIG. 2) when the map of the area desired by the user is displayed on the display 6 will be concretely described using the flow chart shown in FIG. Explained. First, it is assumed that the user wants to see the map of "Sapporo City, Hokkaido", and therefore uttered the vocabulary "Sapporo City, Hokkaido" in step SP3 in FIG. In response to the speech recognition result of this vocabulary, the arithmetic processing unit 4 executes the step S in FIG.
In P6, it is determined that the operation information is map display operation information.

Thereafter, as shown in FIG. 3, the arithmetic processing unit 4 enters from step SP10 and checks in step SP11 whether or not there is a map of the first candidate area in the candidate list which is the result of voice recognition of the vocabulary. If there is no map of the first candidate, the process proceeds to step SP12, where a warning sound is emitted and a warning message such as "No map" is displayed on the display 6 and then the process proceeds to step SP8. To finish. On the other hand, step SP1
If there is the first candidate map in step 1, step S
The process shifts to P13, where the map data S3 is read from the CD-ROM 5 and the map of the first candidate is displayed on the display 6.

Next, the arithmetic processing unit 4 determines the state of the candidate list switch which is a soft switch in step SP14. As a result, if the candidate list switch is not set to the ON state, the operation proceeds to step SP8 and the processing is terminated. If the candidate list switch is set to the ON state in advance, the process proceeds to the next step SP15 and the candidate list of vocabulary that can be expected is displayed on the display 6. For example, as shown in FIG. 4, a candidate list 12 of predictable vocabulary is displayed on the first candidate map 11 in order of probable vocabulary. At that time, since the current state indicates that the first candidate is selected (that is, the map displayed earlier indicates that it is the first candidate), the first candidate portion is highlighted.

Next, the arithmetic processing unit 4 activates a timer for counting the time Ts in step SP16, and determines in the following step SP17 whether or not there is a re-voice by the user. As a result, if there is a re-voice, the timer is ended in step SP18, the process returns to the step SP11 and waits for the recognition result from the voice recognition processing unit 8. If no re-voice is heard, the process proceeds to step SP19. .

In step SP19, the arithmetic processing unit 4
Determines whether or not the remote commander 7 has been operated by the user. If the remote commander 7 has been operated, the timer is terminated at step SP20 and the process proceeds to step SP21. In step SP21, it is determined whether or not the user has selected a new candidate from the candidate list, and if it is not the selection of a new candidate, it is determined in step SP22 whether the "return" command has been selected, and "return" is performed. If it is not the command of step S
Return to P17. If the command is a "return" command, the display of the candidate list is terminated, and then the process proceeds to step SP8 to terminate the process. If the new candidate is selected in step SP21, the process proceeds to step SP23, where the map data S3 is read from the CD-ROM 5 and the map of the area represented by the selected new candidate vocabulary is displayed on the display 6. After that, the process proceeds to step SP8 to end the process.

On the other hand, if the remote commander 7 is not operated by the user in step SP19, the process proceeds to step SP24, where it is determined whether or not the count time of the timer is longer than the time Ts. As a result, if the count time of the timer is not large, the process returns to step SP17 and repeats the process. If the count time is large, the timer is ended in step SP25, and then the display of the candidate list is ended in step SP23 and the first candidate is selected. The map is displayed, and the process is ended by moving to step SP8.

In this manner, the arithmetic processing unit 4 first displays the map of the area represented by the first candidate vocabulary, and displays the candidate list for the time Ts if the candidate list switch is set to the ON state in advance. If a selection operation is performed by the user during the display, a map of the area indicated by the selected new candidate vocabulary is displayed. As a result, even if the first candidate vocabulary is erroneously recognized, the erroneous recognition can be easily corrected.

Incidentally, when performing a selection operation while the candidate list is displayed, the cursor key provided on the remote commander 7 is operated to highlight the portion to be selected, and thereafter, the remote commander 7 is also provided. Operate the Enter key. By this operation, the vocabulary candidate desired by the user or the "return" command can be selected.

In the above configuration, when inputting operation information by voice, the user first turns on the microphone 9 and utters a vocabulary representing the operation information, and sends the voice signal S5 to the voice recognition processing section 8. The voice recognition processing unit 8 recognizes the vocabulary spoken by the user based on the voice signal S5. At that time, the speech recognition processing unit 8 obtains up to the fifth candidate vocabulary that can be predicted as the most probable vocabulary as the first candidate, and outputs the obtained candidate list to the arithmetic processing unit 4 as the recognition result S6.

The arithmetic processing unit 4 discriminates the operation information based on the vocabulary of the first candidate in the candidate list, and performs the processing based on the discriminated operation information. Then, if the candidate list switch is set to the ON state in advance, the arithmetic processing unit 4 displays the candidate list on the display 6 for a predetermined time, and the vocabulary of the second and subsequent candidates is displayed by the user's operation during the display. If selected, the processing based on the selected vocabulary is performed.

As described above, in the case of this embodiment, the processing is first performed based on the most probable first candidate vocabulary among the vocabulary recognized by the voice recognition processing unit 8, and then,
A list of possible vocabulary candidates is displayed so that the user can select and operate the second and subsequent vocabulary words. By displaying the candidate list of possible vocabulary and selecting the vocabulary of the second and subsequent candidates, if the vocabulary of the first candidate is erroneously recognized, the vocabulary can be immediately and easily erroneously recognized. The recognition can be corrected, and the usability can be improved without having to start over from the beginning as in the past.

According to the above configuration, the vocabulary uttered by the user is predicted based on the input voice signal S5, a plurality of predictable vocabularies are obtained, and the obtained plurality of vocabularies are arranged in a probable order. A voice recognition processing unit 8 for creating
First, processing is performed based on the vocabulary of the first candidate in the candidate list, and then the candidate list is displayed so that the vocabulary of the second and subsequent candidates can be selected by the user. When the vocabulary of is selected, by providing the arithmetic processing unit 4 that performs processing based on the vocabulary,
When the vocabulary of the first candidate is erroneously recognized, the erroneous recognition can be corrected immediately and easily, and the usability can be improved as compared with the related art.

In the above embodiment, the case where the candidate list is simply displayed has been described, but the present invention is not limited to this, and when the candidate list includes vocabularies of different types of commands, for example, As shown in FIG. 5, the vocabularies of the different types of commands may be displayed differently by changing the character color so that they can be distinguished.

In the above embodiment, the case where the display 6 is a liquid crystal display has been described. However, the present invention is not limited to this, and may be a CRT (Cathode-Ray Tube) display, for example, and a candidate list or the like. Other display means may be used as long as the information can be displayed.

Further, in the above-mentioned embodiment, the case where the voice recognition processing unit 8 obtains up to the fifth candidate has been described.
The present invention is not limited to this, and the number of candidates may be any.

Further, in the above-mentioned embodiment, the case where the candidate list is displayed when the candidate list switch is set to the ON state has been described, but the present invention is not limited to this, and the re-voice of the user is voiced. The number of times may be counted, and the candidate list may be displayed when the count value exceeds a predetermined number. That is, as in the case above,
First, the arithmetic processing unit 4 performs processing based on the vocabulary of the first candidate. If the first candidate vocabulary is erroneously recognized, the processing is different from the processing desired by the user. As a result, the user calls back to correct the process. The arithmetic processing unit 4 counts the number of re-voices, and if the number of erroneous recognitions is equal to or more than a predetermined number (that is, if the operation is redone many times), the candidate list is displayed and the user is made to perform a selection operation. By doing so, the candidate list can be displayed only when misrecognition occurs.

Further, in the above-mentioned embodiment, the case where the display of the candidate list is controlled by the arithmetic processing unit 4 has been described, but the present invention is not limited to this, and the display of the candidate list can be controlled. Other control means may be used. In short, a control unit that first performs processing based on the vocabulary of the first candidate in the candidate list supplied from the voice recognition processing unit 8, then displays the candidate list, and performs processing based on the vocabulary selected by the user. By providing, the same effect as the above case can be obtained.

In the above embodiment, the case where the present invention is applied to the on-vehicle navigation device 1 has been described, but the present invention is not limited to this, and the operation information can be input by voice. If applicable, it can be applied to other navigation devices, and further widely applied to other electronic devices.

[0039]

As described above, according to the present invention, the vocabulary uttered by the user is predicted based on the input voice signal,
A voice recognition processing means for creating a candidate list in which the expected vocabulary is arranged in a probable order, and a processing based on the vocabulary of the first candidate in the candidate list are performed, and then the candidate list is displayed on the display means to display the second candidate. A vocabulary after the candidate is made selectable by the user, and when a vocabulary after the second candidate is selected by the user, a control means for performing processing based on the selected vocabulary is provided. As a result, when the first candidate vocabulary is erroneously recognized, it is possible to immediately and easily correct the erroneous recognition without redoing it from the beginning as in the past, and thus it is easier to use than the conventional one. Can be improved. Thus, it is possible to realize a voice recognition device and a navigation device using the same, which can improve usability.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a vehicle-mounted navigation device according to an embodiment of the present invention.

FIG. 2 is a flowchart showing a processing procedure when inputting operation information by voice.

FIG. 3 is a flow chart showing a processing procedure of an arithmetic processing unit in the case of displaying a map.

FIG. 4 is a schematic diagram showing a candidate list displayed on a display.

FIG. 5 is a schematic diagram showing a candidate list displayed according to another embodiment.

[Explanation of symbols]

1 ... In-vehicle navigation device, 3 ... GPS receiving unit, 4 ... Arithmetic processing unit, 5 ... CD-ROM, 6 ... Display, 7 ... Remote commander, 8 ... Voice recognition processing unit, 9 ... Mike, 10 ... Talk switch.

Continuation of front page (51) Int.Cl. ⁶ Identification number Office reference number FI Technical display location G10L 3/00 561 G10L 3/00 561E // G08G 1/0969 G08G 1/0969

Claims (10)

[Claims] 1. A vocabulary spoken by a user is predicted based on an input voice signal, and a plurality of predictable vocabularies are obtained,
A speech recognition processing means for creating and outputting a candidate list in which a plurality of obtained vocabularies are arranged in a probable order, and processing based on the first candidate vocabulary in the candidate list output from the speech recognition processing means are performed. After that, the candidate list is displayed on a predetermined display unit so that the user can select the vocabulary after the second candidate, and when the vocabulary after the second candidate is selected by the user, the selection is performed. A voice recognition device, comprising: a control unit that performs a process based on the generated vocabulary. 2. The control means displays the candidate list on the display means for a predetermined time,
If there is no selection operation by the user during the display,
The voice recognition device according to claim 1, wherein the process returns to the process based on the first candidate vocabulary. 3. The voice recognition device according to claim 2, wherein the control means displays the candidate list on the display means for a predetermined time by activating a predetermined timer. 4. The voice recognition apparatus according to claim 1, wherein the control means displays the candidate list if the candidate list display is set in advance by the user. 5. The voice recognition device according to claim 1, wherein the control means counts the number of times the user has re-voiced, and displays the candidate list when the count value reaches a predetermined number or more. 6. A navigation device for receiving a radio wave from a satellite to calculate the position of a moving body, superimposing the calculated position on a map and displaying the result on a predetermined display means, based on an input audio signal. From the speech recognition processing means for predicting the vocabulary uttered by the user, obtaining a plurality of predictable vocabularies, creating a candidate list in which the obtained plurality of vocabularies are arranged in a probable order, and outputting the candidate list. After performing processing based on the vocabulary of the first candidate in the output candidate list, the candidate list is displayed on the display means so that the vocabulary of the second and subsequent candidates can be selected by the user. Second by
A navigation device characterized in that when a vocabulary after a candidate is selected, the control device performs a process based on the selected vocabulary, and is operated by voice. 7. The control means displays the candidate list on the display means for a predetermined time,
If there is no selection operation by the user during the display,
7. The navigation device according to claim 6, wherein the process returns to the process based on the first candidate vocabulary. 8. The navigation device according to claim 7, wherein the control means displays the candidate list on the display means for a predetermined time by activating a predetermined timer. 9. The navigation device according to claim 6, wherein the control means displays the candidate list if the candidate list display is set in advance by the user. 10. The navigation device according to claim 6, wherein the control means counts the number of times the user has re-voiced, and displays the candidate list when the count value reaches a predetermined number or more.