JP2000020092A - Dictation device and recording medium recording dictation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate the possibility that speeches not intended by a user are inputted by attaining the state of not accepting speech input exclusive of a specific command speech when the user interrupts utterance for a specified period of time. SOLUTION: The speech introduced from an input section 11 is subjected to recognition processing and are converted to character strings in a speech recognition section 14. These character string are processed through a command forming section 15 or text forming section 17. At this time, the input intervals of the speech are monitored in a speech input judgment section 17. If the speeches are not inputted even after the specified time, a speech input mode is turned off to make the character strings inputted during this time ineffective. As a result, the possibility that the unnecessary speeches, such as external sounds and the speeches not intended by the user, are inputted while the user interrupts the speech input is eliminated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dictation device used for, for example, a personal computer and for recognizing a word uttered by a user and inputting it as text, particularly when the user interrupts voice input. The present invention relates to a dictation device having a function of controlling reception of unnecessary voice input and a recording medium recording a dictation program.

[0002]

2. Description of the Related Art Along with the advance of speech recognition technology, a dictation system for inputting text by voice has been developed. In a dictation system, text input from a keyboard or the like is input by voice.

Normally, control codes other than text (line feed, line feed,
Delete) can be entered from the keyboard,
A specific command voice (for example, “new line symbol” or “deletion symbol”) registered in advance is input by speaking.

Here, in the conventional dictation system, on / off switching of the voice input mode is often controlled from a keyboard, a mouse, or the like. Speech is recognized. Therefore, in addition to the voice uttered intentionally by the user, for example, a voice coming from the outside or a voice uttered without intention of input by the user (such as muttering or groaning) may be input.

Conventionally, when a user interrupts voice input for a long period of time, for example, when the user leaves a seat or stops inputting, the voice input mode is turned off through a keyboard or the like, and unnecessary voice is input. Danger was avoided.

[0006]

As described above, in order to avoid unnecessary voice input, when the voice input is interrupted for a long period of time, the voice input mode may be turned off. However, rather than a long break,
When the user temporarily suspends for a short period of time, for example, considering the next input, it is necessary to instruct the voice input off mode from the keyboard or the like every time, which is very troublesome.

[0007] In addition, the short-term interruption frequently occurs during voice input, and is also the interval at which a voice that the user does not intend to input (such as muttering) is most likely to be uttered. For this reason, the conventional method has a disadvantage that the possibility of noise being input during voice input cannot be completely excluded.

[0008] The present invention has been made in view of the above points, and when the user interrupts the utterance for a certain period of time, by excluding a specific command voice, the user is prevented from accepting voice input, thereby allowing the user to make an intention. It is an object of the present invention to provide a dictation device and a recording medium on which a dictation program is recorded, which eliminates the possibility of inputting undesired voice.

[0009]

SUMMARY OF THE INVENTION According to the present invention, when a text is created by recognizing a voice uttered by a user, a voice input interval is constantly monitored, and when a voice is interrupted for a predetermined time or longer, a voice input is performed. Turn off the mode and invalidate the character string input during that time, or enter a state in which no input is accepted except for a specific character string, for example, a command for starting the system. .

According to such a configuration, in the dictation system, when the voice input is interrupted for a certain time,
The voice input mode is automatically turned off, so that no unnecessary voice input is accepted. Therefore, it is possible to eliminate a possibility that an unnecessary sound such as an external sound or a sound not intended by the user is input while the user stops the sound input.

[0011]

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

(First Embodiment) FIG. 1 is a block diagram showing a configuration of a dictation device according to a first embodiment of the present invention. The present device recognizes words spoken by a user and inputs the words as text.
For example, it is realized by a computer which reads a program recorded on a recording medium such as a magnetic disk and the operation of which is controlled by the program.

As shown in FIG. 1, the present apparatus comprises an input unit 1
1, control unit 12, display unit 13, voice recognition unit 14, command generation unit 15, text generation unit 16, voice input determination unit 1
7, a command dictionary 18, and a text storage unit 19.

The input section 11 is for inputting a voice uttered by the user, and is composed of an input device such as a microphone.

The control unit 12 controls the entire apparatus, and is, for example, a central processing unit (CPU), and each processing unit (speech recognition unit 14, command generation unit 15, text generation unit) for executing a text creation process. 16. Voice input determination unit 1
The control of 7) is performed. Further, the control unit 12 is provided with a mode storage unit 12a for storing the ON / OFF state of the voice input mode. When the voice input mode is on, the information input by voice is valid, and when the voice input mode is off, the information input by voice is invalid.

The display unit 13 is for displaying a character string of text converted by voice recognition, and is composed of, for example, a CRT display device or a liquid crystal display device (a flat panel display device such as a liquid crystal display device).

The voice recognition unit 14, command generation unit 15, text generation unit 16, voice input determination unit 17, command dictionary 18, and text generation unit 19 are functional elements required for text creation processing. (And a CPU that executes the processing routine).

The voice recognition unit 14 performs a recognition process on the voice input from the input unit 11 with reference to the voice recognition dictionary 14a, converts the voice into a character string obtained as a result of the recognition, and outputs the character string. Note that the voice recognition processing (recognition engine) by the voice recognition unit 14 is always activated regardless of whether the voice input mode is on or off.

The command generation unit 15 searches whether the character string converted by the voice recognition unit 14 is registered in the command dictionary 18. Then, if registered, the control code of the corresponding command is output, and if not registered, the character string is output as it is.

The text generation unit 16 stores the character string passed from the command generation unit 15 in the text storage unit 19 when the voice input mode is on, and stores the text string in the text storage unit 19 when the voice input mode is off. Do not store.

The voice input judging section 17 monitors the interval between voice inputs, and if no voice is input after a certain period of time when the voice input mode is on, the mode storage section 1 of the control section 12.
The voice input mode set to 2a is turned off.

The command dictionary 18 holds a combination of a character string representing a command, a corresponding control code, and a command execution level. FIG. 4 shows an example of the command dictionary 18.

In the example shown in FIG. 4, a character string "ONSENRYOKYODOON" is associated with a command "turn on voice input mode", and its level is set to "1". ing. Similarly, the character string "Onsei-ryokumodofu" is associated with the command "Turn off the voice input mode", and its level is set to "1". In addition, the character string “kaigyo” is associated with the command “line feed”, and its level is set to “0”. Similarly, the character string "kaipaiji" is associated with the command "page break",
The level is set to “0”.

The former command is directly related to the control of the system itself, and is registered as level "1". The latter command relates to document editing and does not directly relate to control of the system itself.
Therefore, this type of command is registered as level “0”.

The text storage unit 19 stores a text character string converted by voice recognition.

Next, the operation of the first embodiment will be described.

First, when a voice is input from the input unit 11, the voice is sent to the voice recognition unit 14 and the voice input determination unit 17.
The voice recognition unit 14 recognizes the input voice, converts the voice into a word character string, and sends the word character string to the command generation unit 15.

The command generation unit 15 searches for a command character string registered in advance in the command dictionary 18 by using the word character string as a key. It is checked whether the level of the command to be executed corresponds to a preset level. As a result, if applicable, the command processing corresponding to the character string is executed. If not, the word character string is sent to the text generator 16 as it is.

The text generator 16 stores the sent word character string in the text storage 19 when the voice input mode is on, and stores the text in the text storage 19 when the voice input mode is off. do not do. The text stored in the text storage unit 19 is constantly displayed on the display unit 13.

Here, the voice input mode is turned on in the flow from the input unit 11 to the command generation unit 15 described above. When a command voice for turning on the voice input mode is input from the input unit 11, the command generation unit 15 confirms that the command character string exists in the command dictionary 18, and outputs a command corresponding to the command character string. Output control code.

The voice input determination unit 17 monitors the interval between voice inputs. When the voice input mode is turned on, if the interval between voices sent from the input unit 11 exceeds a certain time, voice input is determined. Turn off mode.

Next, a detailed description will be given with reference to the flowchart of FIG.

FIG. 2 is a flowchart showing the operation of the voice processing in the first embodiment. When a voice is input from the input unit 11, the control unit 12 transmits the voice to the voice recognition unit 14.
(Step A11). The voice recognition unit 14 uses the voice recognition dictionary 14a to perform a recognition process on the input voice to convert the voice into a word character string, and passes the word character string to the command generation unit 15 (step A12).

The command generation unit 15 includes a voice recognition unit 14
It is checked whether the word character string obtained by the above matches the command character string registered in the command dictionary 18 (step A13). Then, when the character string does not match the command character string, that is, when a character string other than the command is input (No in step A13), the command generation unit 15 checks the current state of the voice input mode (step A14).

As a result, if the voice input mode is on (Yes in step A14), the command generation unit 15 passes the word character string at that time to the text generation unit 16 and stores it in the text storage unit 9 (step A15). ). When the voice input mode is off (N in step A14)
o), the text is not stored in the text storage unit 19, and the state is in a state of waiting for voice input.

More specifically, for example, it is assumed that a voice such as “Introduction” is input from the input unit 11 while the voice input mode is on. This input voice is passed to the voice recognition unit 14 and is converted into a word character string such as “Introduction”. The command generator 15 searches for a corresponding command character string from the command dictionary 18 as shown in FIG.
In this case, since there is no matching command character string, it is sent to the text generation unit 16 as it is. Command generator 15
Is stored in the text storage unit 19.

On the other hand, if the word character string matches the command character string registered in the command dictionary 18 in step A13, the command generator 15 checks whether the voice input mode is on (step A16). When the voice input mode is off (N in step A16)
o), the command generation unit 15 checks whether or not the level of the command character string that matches the word character string corresponds to a preset level (step A17).

As a result, if the set level is not satisfied,
That is, when the command character string is a low-level command character string used for document editing (No in step A17),
Invalidates the input of the character string and does not execute the corresponding command processing. Also, if it corresponds to the setting level,
That is, when the command character string is a high-level command character string related to the control of the system itself (Yes in step A17).
s) The input of the character string is validated, and the corresponding command processing is executed (step A18).

In step A16, when the voice input mode is ON, regardless of the command level, the input of the command character string at that time is validated, and the command processing corresponding thereto is executed (step A1).
8).

For example, it is assumed that after the voice of "Introduction" is input, the next voice is not input for a certain period of time and the voice input mode is turned off.

In this state, it is assumed that a voice "line feed" is input. The input voice “Kaigayo” is converted into a word character string “Kaigayo” by the voice recognition unit 14 and compared with a command character string registered in the command dictionary 18. In this case, the word character string “Kaikyo” is
As shown in FIG. 7, the command character string matches the command character string registered in the command dictionary 18.

In step A16, since the voice input mode is in the OFF state, it is checked in the command dictionary 18 whether the level of "OK" corresponds to a preset command level. In this case, it is assumed that “only level 1 commands are accepted” is set. The level of “Kaikyo” is “0” as shown in FIG.
Therefore, the corresponding command processing is not executed.

Subsequently, it is assumed that a voice "voice input ON" is input. This input voice is converted into a word character string “ONSENRYOKURIDOON” by the voice recognition unit 14 and compared with a command character string registered in the command dictionary 18. Since the voice input mode is off, the level of "ONSENRYUKRO-DON" is checked by the command dictionary 18 in step A17. In this case, the level of the command "ONSEI RYOUKUMO DON" is "1", which corresponds to the level of the command set in advance. Therefore, in step A18, the processing of the corresponding command "turn on the voice input mode" is executed.

Here, the preset command level may be set such as "accept only commands of level 0" or "accept commands of all levels".

Next, the mechanism for turning off the voice input mode will be described with reference to the flowchart of FIG.

FIG. 3 is a flowchart showing the operation of the voice input determination process in the first embodiment. The voice input determining unit 17 constantly monitors the state of the voice input mode (step B11). When the voice input mode is turned on (Yes in step B11), the voice input determination unit 1
7 checks the input interval of the sound passed from the input unit 11 (step B12). Here, in the case where no voice is input for a certain period of time (step B12).
Yes), the voice input determination unit 17 turns off the voice input mode state set in the mode storage unit 12a of the control unit 12 (step B13).

For example, it is assumed that after a predetermined time has elapsed since the voice of "Introduction" was input, the voice of "Em" was input. In such a case, the voice input determination unit 1
Reference numeral 6 monitors the interval between voice inputs while determining that the voice input mode is on, confirms that no voice has been input for a predetermined time after the voice of "Introduction", and turns off the voice input mode.

Therefore, the voice "Eto" input from the input unit 11 is converted to the word character string "Eto" by the voice recognition unit 14.
When the character string is converted to the character string, the text generation unit 16 determines that the voice input mode is off, so that the character string is not stored in the text storage unit 19. That is, the text storage unit 19 stores only a character string such as “Introduction” input before the voice input mode is turned off. As a result, as shown in FIG. 5, the display unit 13 displays the text “Introduction” stored in the text storage unit 19.

As described above, by turning off the voice input mode when no voice is input even after a certain period of time has elapsed, the external sound (noise) generated between voice inputs can be obtained.
Also, it is possible to avoid inputting unnecessary voice as text, such as a voice unintentionally uttered by the user. If a specific command is input by voice during this time, the processing of that command eliminates unnecessary voice input and allows voice input directly related to system control itself to be handled as it is. it can.

(Second Embodiment) Next, a second embodiment of the present invention will be described.

FIG. 6 is a block diagram showing the configuration of the dictation device according to the second embodiment of the present invention. This apparatus is also realized by a computer which reads a program recorded on a recording medium such as a magnetic disk and the operation of which is controlled by this program, similarly to FIG. Is the same as

6 differs from FIG. 1 in that a voice input mode warning unit 20 is provided.
The voice input mode warning unit 20 includes a voice input determination unit 17.
Is performed on the display unit 13 to display a message indicating that the voice input mode is on or off in accordance with the determination result.

Next, the operation of the second embodiment will be described.

The operation of the configuration of FIG. 6 is exactly the same as the operation of the configuration of FIG. 1 in the first embodiment.
The details of the operation of the audio processing according to the flowchart in FIG. The difference is that a process of the voice input mode warning unit 20 is added to the mechanism for turning off the voice input mode. This will be described with reference to the flowchart of FIG.

FIG. 7 is a flowchart showing the operation of the voice input determination process in the second embodiment. The voice input determination unit 17 constantly monitors the state of the voice input mode (step C11). When the voice input mode is turned on (Yes in step C11), the voice input determination unit 1
7 checks the input interval of the sound passed from the input unit 11 (step C12). Here, when no voice is input for a certain period of time (step C12).
Yes), the voice input determination unit 17 turns off the voice input mode state set in the mode storage unit 12a of the control unit 12 (step C13).

When the voice input mode is turned off, the voice input determination unit 17 activates the voice input mode warning unit 20 to notify that the voice input mode has been turned off. In response to this, the voice input mode warning unit 20 displays on the display unit 13 that the voice input mode is off (step C1).
4).

Here, it is assumed that the voice input mode is turned on in a state where “Introduction” is stored in the text storage unit 19. Since the voice input mode is on, the voice input determination unit 16 checks the interval between voices. Then, if the voice following “Introduction” is not input even after the elapse of a certain time, the voice input determination unit 16 turns off the voice input mode and notifies the voice input mode warning unit 20.
The voice input mode warning unit 20 notifies the user that the voice input mode has been turned off by displaying a message 31 on the display unit 13 as shown in FIG.

As described above, when the voice input mode is automatically turned off, the fact is notified to the user, so that it is possible to prevent the user from making useless utterances without noticing the mode off state. .

In the present embodiment, the user is notified that the voice input mode has been turned off by displaying a message 31 as shown in FIG. 8, but other notification methods include, for example, a warning lamp. Lighting, generation of a warning sound, and further, a voice message such as "The voice input mode has been turned off."

(Third Embodiment) Next, a third embodiment of the present invention will be described.

FIG. 9 is a block diagram showing the configuration of a dictation device according to the third embodiment of the present invention. This apparatus is also realized by a computer which reads a program recorded on a recording medium such as a magnetic disk and the operation of which is controlled by this program, similarly to FIG. Is the same as

The configuration of FIG. 9 differs from that of FIG. 1 in that a voice input mode switching unit 21 is provided. The voice input mode switching unit 21 performs a process of switching the voice input mode on when a command character string is input while the voice input mode is off.

Next, the operation of the third embodiment will be described.

First, when a voice is input from the input unit 11, the voice is sent to the voice recognition unit 14 and the voice input determination unit 17.
The voice recognition unit 14 recognizes the input voice, converts the voice into a word character string, and sends the word character string to the command generation unit 15.

The command generation unit 15 searches for a command character string registered in advance in the command dictionary 18 with the word character string, and if there is a corresponding command character string, the command generation unit 15 generates a command corresponding to the character string. Check whether the level corresponds to a preset level. As a result, if applicable, the corresponding command processing is executed. At this time, the voice input mode switching unit 21 is activated. If not, the word character string is used as it is in the text generator 1
Send to 6.

The voice input mode switching section 21 checks whether or not the voice input mode is on, and if it is off, turns it on.

When the voice input mode is on, the text generation unit 16 stores the sent word character string in the text storage unit 19. When the voice input mode is off, the data is not stored in the text storage unit 19. Text storage unit 19
Are displayed on the display unit 13 constantly.

Turning on the voice input mode is performed in the flow from the input unit 11 to the command generation unit 15 described above. When a command voice for turning on the voice input mode is input from the input unit 11, the command generation unit 15 confirms that the command character string exists in the command dictionary 18, and outputs a command corresponding to the command character string. Output control code.

The voice input judging section 17 monitors the interval between voice inputs. When the voice input mode is turned on, if the interval between voices sent from the input section 11 exceeds a predetermined time, voice input is determined. Turn off mode.

Next, a detailed description will be given with reference to the flowchart of FIG.

FIG. 10 is a flowchart showing the operation of the voice processing in the third embodiment. When a voice is input from the input unit 11, the control unit 12 transmits the voice to the voice recognition unit 1.
4 (step D11). The voice recognition unit 14 uses the voice recognition dictionary 14a to perform a recognition process on the input voice, converts the voice into a word character string, and passes the word character string to the command generation unit 15 (step D12).

The command generation unit 15 includes a voice recognition unit 14
It is checked whether or not the word character string obtained by the above matches the command character string registered in the command dictionary 18 (step D13). When it does not match the command character string, that is, when a character string other than the command is input (No in step D13), the command generation unit 15 checks the current state of the voice input mode (step D14).

As a result, if the voice input mode is on (Yes in step D14), the command generation section 15 passes the word character string at that time to the text generation section 16 and stores it in the text storage section 9 (step D15). ). When the voice input mode is off (N in step D14)
o), the text is not stored in the text storage unit 19, and the state is in a state of waiting for voice input.

On the other hand, if the word character string matches the command character string registered in the command dictionary 18 in step D13, the command generation unit 15 checks whether the voice input mode is on (step D16). If the voice input mode is off (N in step D16)
o), the command generation unit 15 checks whether or not the level of the command character string that matches the word character string corresponds to a preset level (step D17).

As a result, if the command character string level does not correspond to the set level (N in step D17)
o) The input of the character string is invalidated, and the corresponding command processing is not executed. In addition, when the input level corresponds to the set level (Yes in step D17), the input of the character string is validated, and the corresponding command processing is executed (step D18).

Here, the command generation unit 15 activates the voice input mode switching unit 21. The voice input mode switching unit 21 checks whether the voice input mode is on or off, and if it is off, switches the voice input mode on (step D19).

When the voice input mode is on (Yes in step D16), the command character string at that time is validated regardless of the command level, and the processing corresponding to the command character string is executed (step D20). ).

For example, when the voice input mode is off,
Suppose that a voice "page break" is input. The voice of “page break” is converted by the voice recognition unit 14 into a word character string of “kai page”. When the command generator 18 searches the command dictionary 18, a matching command character string exists.

When the voice input mode is checked in step D16, since the voice input mode is off, it is checked in the command dictionary 18 whether the level of the command character string "kai page" corresponds to the level of a preset command. .

Here, if "all levels of commands are accepted" is set, the command character string "kai page" is a command of the set level, and the corresponding command "command" is set at step D19. Page break "is executed.

Further, the command input unit 15 activates the voice input mode switching unit 21. The voice input mode switching unit 21 checks whether the voice input mode is on. In this case, since the voice input mode is in the off state, it is switched to the on state.

Here, the preset command level may be set such as "accept only level 0 commands" or "accept only level 1 commands".

As described above, when the input voice is the voice for the command, it is determined that the user has resumed the interrupted voice input, and the state is switched to a state in which the voice input is accepted. Thus, when the voice input is not accepted (the voice input mode is off), when the user resumes the voice input, the voice for the command is issued without giving a command to change the status from a keyboard or the like. Immediately after the input, voice input can be performed.

(Fourth Embodiment) Next, a fourth embodiment of the present invention will be described.

FIG. 11 is a block diagram showing a configuration of a dictation device according to a fourth embodiment of the present invention. This apparatus is also realized by a computer which reads a program recorded on a recording medium such as a magnetic disk and the operation of which is controlled by this program, similarly to FIG. Is the same as

The configuration of FIG. 11 differs from that of FIG. 1 in that a recognition rate determination unit 22 is provided. The recognition rate determination unit 22 determines whether the recognition rate of the word character string recognized by the speech recognition unit 14 is higher than a certain value.

Next, the operation of the fourth embodiment will be described.

First, when a voice is input from the input unit 11, the voice is sent to the voice recognition unit 14 and the voice input determination unit 17.
The voice recognition unit 14 recognizes the input voice, converts the voice into a word character string, and sends the word character string to the command generation unit 15.

The command generation unit 15 searches for a command character string registered in advance in the command dictionary 18 using the word character string, and if there is a corresponding command character string, the command generation unit 15 searches for a command corresponding to the character string. Check whether the level corresponds to a preset level. As a result, if applicable, the corresponding command processing is executed. If not, an inquiry is made to the recognition rate determination unit 22.

The recognition rate determining section 22 checks whether or not the recognition rate of the word character string recognized by the voice recognition section 14 exceeds a preset value. If so,
The voice input mode is turned on, and control is returned to the command generation unit 15. The command generation unit 15 sends the word character string to the text generation unit 16.

The text generator 16 stores the sent word character string in the text storage 19 when the voice input mode is on, and stores it in the text storage 19 when the voice input mode is off. Do not. The text stored in the text storage unit 19 is constantly displayed on the display unit 13.

Turning on the voice input mode is performed by the flow from the input unit 11 to the command generation unit 15 described above. When a command voice for turning on the voice input mode is input from the input unit 11, the command generation unit 15 confirms that the command character string exists in the command dictionary 18, and outputs a command corresponding to the command character string. Output control code.

The voice input judging section 17 monitors the interval between voice inputs. When the voice input mode is turned on, if the interval between voices sent from the input section 11 exceeds a certain time, voice input is determined. Turn off mode.

Next, a detailed description will be given with reference to the flowchart of FIG.

FIG. 12 is a flowchart showing the operation of the voice processing in the fourth embodiment. When a voice is input from the input unit 11, the control unit 12 transmits the voice to the voice recognition unit 1.
4 (step E11). The voice recognition unit 14 uses the voice recognition dictionary 14a to perform a recognition process on the input voice to convert the voice into a word character string, and passes the word character string to the command generation unit 15 (step E12).

In the command generation unit 15, the voice recognition unit 14
It is checked whether the word character string obtained by the above matches the command character string registered in the command dictionary 18 (step E13). If it matches the command character string, the command generator 15 checks whether the voice input mode is on (step E14).

If the voice input mode is off (No in step E14), the command generator 15 checks whether the level of the command character string that matches the word character string corresponds to a preset level. (Step E1
5).

As a result, if it does not correspond to the set level,
That is, if the command character string is a low-level command character string used for document editing (No in step E15),
Invalidates the input of the character string and does not execute the corresponding command processing. Also, if it corresponds to the setting level,
That is, when the command character string is a high-level command character string relating to the control of the system itself (Ye in step E15).
s) The input of the character string is validated, and the corresponding command processing is executed (step E16).

On the other hand, if the word character string does not match the command character string registered in the command dictionary 18 at the step E13, an inquiry is made to the recognition rate judgment section 22.

For example, when the voice input mode is off,
It is assumed that a voice saying “End” is input from the input unit 11. The voice of “End” is passed to the voice recognition unit 14 and is converted into a word character string of “End”. The command generation unit 15 searches the command dictionary 18 but inquires of the recognition rate determination unit 22 because there is no corresponding command character string.

The recognition rate determination section 22 determines whether or not the recognition rate of the word character string obtained by the voice recognition section 14 is larger than a preset value (step E17). As a result, when the recognition rate of the word character string exceeds the set value (Yes in step E17), the voice input mode is turned on (step E18), and the control is returned to the command generation unit 15. The command generator 15 passes the word character string to the text generator 16 as it is.

In this case, the speech recognition dictionary 14a provided in the speech recognition section 14 has registered therein information for speech recognition of a word relating to text creation in advance. Therefore, when a word related to text creation is input by speech, the word can be recognized at a high recognition rate, and can be distinguished from noise.

Here, it is assumed that the word character string "conclusion" is not a noise but a correct word, so that the recognition rate is high.
Then, it is determined that the recognition rate determination rate 12 is higher than the set value, and the voice input mode is turned on.

The text generator 16 checks whether the voice input mode is on (step E19). If the voice input mode is on (Yes in step E19)
s) The word character string passed from the command generation unit 15 is stored in the text storage unit 19 (step E20). If the voice input mode is off (No in step E19),
The process returns to step E11 without storing in the text storage unit 19.

In the above example, the word character string “End” passed from the command generation unit 15 is
Since the voice input mode is turned on by the determination of 2, the data is stored in the text storage unit 19 and displayed on the display unit 13.

As described above, whether or not the input voice is noise is distinguished by the recognition rate of voice recognition, and if the recognition rate is high, it can be determined that the voice input is intended by the user. Therefore, in the state where voice input is not accepted (the voice input mode is off), when the user resumes voice input, the voice input is performed immediately without giving a command to change the status from the keyboard or the like. Will be able to

The present invention is not limited to the above embodiment.

For example, a voice input determining unit 17 for turning off the voice input mode, a recognition rate determining unit 22 for turning on the voice input mode when the recognition rate is larger than a certain value, and a voice for command being input. Although the audio input mode switching unit 21 that turns on the audio mode when the processing is performed is configured to perform the processing separately, they may be performed simultaneously.

In short, the present invention can be variously modified and implemented without departing from the gist.

Further, the methods described in the above-described embodiments include, for example, a magnetic disk (floppy disk, hard disk, etc.), an optical disk (CD-ROM,
It is also possible to write the data on a recording medium such as a DVD or a semiconductor memory and apply it to various devices, or to transmit it via a communication medium and apply it to various devices. A computer that realizes the present apparatus reads the program recorded on the recording medium, and executes the above-described processing by controlling the operation of the program.

[0111]

As described above, according to the present invention, when a text is created by recognizing a voice uttered by a user, the input interval of the voice is constantly monitored, and when the voice is interrupted for a predetermined time or more, By turning off the voice input mode, invalidating the character string input during that time, or by not accepting input except for a specific character string, for example, a command to start the system, While the user interrupts the voice input, it is possible to eliminate a possibility that unnecessary voices such as an external sound and a voice not intended by the user are input.

When the voice input is not accepted (the voice input mode is off), the user is notified of this fact, so that the user does not notice the mode off state and makes useless utterances. Can be avoided.

When a voice for a command is input, it is determined that the user has resumed the interrupted voice input, and the state is switched to a state in which the voice input is accepted. (When the voice input mode is off), when the user resumes voice input, voice input can be performed immediately after voice input for a command without giving a command to change the status from a keyboard or the like. it can. Also, whether or not the input voice is noise is distinguished by the recognition rate of voice recognition. If the recognition rate is high, it is determined that the voice input is intended by the user, so that the voice input is not accepted (voice Input mode is off) and when the user resumes voice input,
Voice input can be performed immediately without giving a command to change the state from a keyboard or the like.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a dictation device according to a first embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of audio processing in the first embodiment.

FIG. 3 is a flowchart illustrating an operation of a voice input determination process according to the first embodiment.

FIG. 4 is a diagram showing a configuration of a command word in the first embodiment.

FIG. 5 is a view showing a display screen in the first embodiment.

FIG. 6 is a block diagram showing a configuration of a dictation device according to a second embodiment of the present invention.

FIG. 7 is a flowchart for explaining the operation of a voice input determination process in the second embodiment.

FIG. 8 is a diagram showing a display screen in the second embodiment.

FIG. 9 is a block diagram showing a configuration of a dictation device according to a third embodiment of the present invention.

FIG. 10 is a flowchart for explaining the operation of audio processing in the third embodiment.

FIG. 11 is a block diagram showing a configuration of a dictation device according to a fourth embodiment of the present invention.

FIG. 12 is a flowchart for explaining the operation of audio processing in the fourth embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... Input part 12 ... Control part 12a ... Mode storage part 13 ... Display part 14 ... Speech recognition part 14a ... Speech recognition dictionary 15 ... Command generation part 16 ... Text generation part 17 ... Speech input judgment part 18 ... Command dictionary 19 ... Text Storage unit 20: voice input mode warning unit 21: voice input mode switching unit 22: recognition rate determination unit

Claims (7)

[Claims] A voice input unit for inputting a voice; a voice recognition unit for recognizing a voice input by the voice input unit to convert the voice into a character string; and a character obtained as a recognition result by the voice recognition unit. Processing means for processing the sequence; monitoring a voice input interval; if no voice is input after a certain period of time from the voice input means, the voice input mode is turned off, and characters input during that time are turned off. A dictation device comprising input control means for invalidating a column. 2. A voice input unit for inputting voice, a voice recognition unit for recognizing a voice input by the voice input unit and converting the voice into a character string, and a character obtained as a recognition result by the voice recognition unit. Processing means for processing a sequence, storage means for storing a specific character string to receive an input, and monitoring of a voice input interval, and when no voice is input even after a certain period of time from the voice input means, A dictation device comprising: an input control unit that turns off a voice input mode and invalidates a character string other than a specific character string stored in the storage unit during the input mode. 3. The dictation device according to claim 1, further comprising a notification unit for notifying that the voice input mode is in an off state. 4. The dictation device according to claim 1, wherein the input control means switches the voice input mode to an on state when a character string relating to a command is input. 5. The input control means switches the voice input mode to an on state when a character string whose recognition rate by the voice recognition means is higher than a preset value is input. The dictation device according to claim 1. 6. A recording medium storing a dictation program for recognizing voice and inputting it as text, comprising: a step of inputting voice; and a step of recognizing the input voice and converting the input voice into a character string. , A procedure for processing a character string obtained as a recognition result, and monitoring a voice input interval.
A computer-readable recording medium that records a program for causing a computer to execute a procedure for invalidating a character string input during that time. 7. A recording medium on which a dictation program for recognizing voice and inputting it as text is recorded, wherein a procedure for inputting voice, a procedure for recognizing the input voice and converting it into a character string are provided. , A procedure for processing a character string obtained as a recognition result, and monitoring a voice input interval.
A computer-readable recording medium which records a program for causing a computer to execute a procedure for invalidating a character string other than a specific character string input during that time.