JP2000260170A - Speech input device and speech recognition system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible to easily recognize a speech to be in an enrolling mode by displaying as prescribed presenting that enrollment is being operated on a display means, when the speech is inputted for making an external device operate data registration for improvement in accuracy of speech recognition processing. SOLUTION: When a SEL-button is inputted two times and inputted further, PC-LINK processing is performed. In an enrollment mode as one of PC-LINK modes, an operator inputs a speech via a microphone and transmits this speech data to a PC. At the same time, to let the operator know that the speech is being inputted, a counter is operated to increment a counter in a display area of a display part 14. In the enrollment mode, the operator can easily recognize that the speech is being inputted in the enrollment mode, by displaying 'EN ROLL' and then operating the counter to increment the counter.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice input device and a voice recognition system.

[0002]

2. Description of the Related Art Voice input is performed using a voice input device such as a digital recorder, and is transferred to an external device such as a personal computer to perform enrollment processing. 2. Description of the Related Art A speech recognition system that performs speech recognition processing using speech recognition software and a process of registering speech characteristics data has been known.

[0003]

However, in the conventional voice recognition system, the voice input device is not provided with a dedicated display function for indicating that the enrolling operation is being performed. It could not be easily recognized.

The present invention has been made in view of such a problem, and an object thereof is to provide a voice input device and a voice recognition system capable of easily recognizing an enroll mode. Is to do.

[0005]

In order to achieve the above object, a voice input device according to a first aspect of the present invention is a voice input device capable of recording independently and capable of data communication with an external device. When a voice is input to the display means and an external device to register data for improving the accuracy of the voice recognition processing, the display means performs a predetermined display indicating that the enrolling operation is being performed. Control means.

A voice recognition system according to a second aspect of the present invention includes a voice input device including a display unit, which can be independently recorded and can perform data communication with an external device, and a voice data transmitted from the voice input device. And a voice recognition processing device for performing voice recognition processing on the received voice data, wherein the voice recognition processing device has an enroll function for registering data for improving the accuracy of voice recognition processing. While the voice is input to the voice input device, the input voice information is transmitted to the voice recognition processing device, and while the enrollment operation is performed in the voice recognition processing device, the voice input is performed. A predetermined display indicating that the enrolling operation is being performed is displayed on the display means of the apparatus.

[0007] Further, a voice input device according to a third aspect of the present invention comprises:
In the voice input device according to the first invention, the display indicating that the enrolling operation is being performed is performed by operating a counter.

[0008] That is, in the first invention, a voice input device capable of recording independently and capable of data communication with an external device is used, and the external device registers data for improving the accuracy of voice recognition processing. When a voice is input to the device, a predetermined display indicating that the enrolling operation is being performed is displayed on the display means.

According to the second aspect of the present invention, there is provided a voice input device provided with a display means, capable of recording data independently and communicating with an external device, and transmitting voice data received from the voice input device in voice. A speech recognition system is configured with the speech recognition processing device for performing the recognition process, and while the enrollment operation for registering data for improving the accuracy of the speech recognition process for the received speech information is performed by the speech recognition processing device, The display means of the voice input device is caused to perform a predetermined display indicating that the enrolling operation is being performed.

[0010] The voice input device according to a third aspect of the present invention includes:
In the voice input device according to the first invention, a display indicating that the enrolling operation is being performed is performed by operating a count.

[0011]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a diagram showing a configuration of a digital recorder to which a voice input device according to the present invention is applied. In FIG. 1, a microphone 1 is connected to a low-pass filter (LPF) 3 via an amplifier (AMP) 2. The low-pass filter 3 is connected to an audio output terminal 18 via an amplifier 17 and a digital signal processor (DS) via an analog / digital converter (A / D) 4.
P) 5 is connected to the T1 terminal. T of this DSP5
A digital-analog converter (D / A) 9, a low-pass filter (LPF) 10, and an amplifier (AMP) 1 are connected to two terminals.
1 is connected to the speaker 12.

The T3 terminal of the DSP 5 is connected to the system control unit (C
PU) 6 is connected to the T4 terminal. T5 of CPU6
A PC connection terminal 16 is connected to the terminal, a display unit (display means) 14 is connected to the T6 terminal via the drive circuit 13, a power control unit 15 is connected to the T7 terminal, and a built-in is built in the T8 terminal. Recording medium 7 is connected. This recording medium 7 may be removable.

Further, the CPU 6 is connected to an operation input unit 8 having various operation buttons (recording button REC, reproduction button PLAY, stop button STOP, skip button SKIP, select button SEL).

The digital recorder of this embodiment is connected to an external device or a personal computer (hereinafter, referred to as a PC) as a speech recognition processing device to constitute a speech recognition system, and performs various processes such as enroll processing and speech recognition processing. A PC-LINK mode is provided for causing the PC to perform processing. Here, it is assumed that enroll work is performed using a digital recorder as a remote control microphone.
That is, first, the audio output terminal 18 is connected to the audio input terminal of the PC by a cable or the like, and then the enroll mode is set. Next, by reading out the text for enrollment displayed on the screen of the PC, the microphone 1 is read.
After the audio is input from the amplifier 2 and amplified by the amplifier 2, unnecessary components are removed by the low-pass filter 3. Next, the signal is amplified to a transmittable signal level by the amplifier 17 and transmitted as an analog signal from the audio output terminal 18 to the PC. Note that a wireless method such as infrared rays may be used.

Also, instead of transmitting the signal to the PC in the form of an analog signal, the sound input from the microphone 1 is converted into a digital signal by the analog-to-digital converter 4 and then compressed by the digital signal processing unit 5 to generate encoded sound data. PC
May be transmitted. In this case, it is assumed that the PC has a function of decoding the encoded audio data received.

The general recording / reproducing operation of the above-mentioned digital recorder will be described below. When the operator presses the REC button, a recording mode is set, and the sound input from the microphone 1 is converted into an electric signal, which is then amplified by the amplifier 2, and
The unnecessary components are removed by F3. Thereafter, the signal is converted into a digital signal by the analog-digital converter 4 and input to the DSP 5. After the DSP 5 performs a compression process on the digital audio signal, the digital audio signal is recorded on the recording medium 7 as audio data under the control of the CPU 6.

When the operator presses the PLAY button, the reproduction mode is set. After the audio data is read from the recording medium 7 under the control of the CPU 6, the DSP 5 performs an expansion process. The expanded audio signal is converted into an analog signal by a digital-to-analog converter 9, its unnecessary components are removed by a low-pass filter 10, amplified by an amplifier 11, and then output as audio from a speaker 12.

If the STOP button is pressed during the above-described sound recording or reproduction operation, the operation at that time is stopped. Various kinds of information are displayed on the display unit 14 according to the mode. The power control unit 15 controls power supplied to each unit in the apparatus and controls power saving.

FIG. 2 is a view showing the external appearance of the above-mentioned digital recorder.
In addition to the Y button, STOP button, SKIP button, and SEL button, a microphone 1, a display unit 14, a speaker 12, a PC connection terminal 16,
An audio output terminal 18 is arranged at a predetermined position.

FIG. 3 shows C based on an operation input from the operator.
It is a flowchart for demonstrating the detail of a process of PU. This flow starts when a battery is loaded. First, in step S1, initial settings for an operation such as a memory check are performed, and then the recorder mode is set, and FIG.
Is displayed on the display unit 14 (step S2). “N” displayed in the display area A of the display unit 14
o. "1" indicates the file number, and "0001" displayed in the display area B indicates the count value of the counter.The count value of the counter is 0 for each file.
To n. “REC” displayed in the display area C is a mode display, and indicates that the current mode is the recording mode. “PLAY” is displayed in the reproduction mode, and nothing is displayed in the stop mode.

Next, it is determined whether or not the SEL button has been input (step S3). If NO, the process proceeds to step S4 to determine whether or not the REC button has been input. If YES here, the process shifts to step S8 to perform the recording process (R
EC processing) and returns to step S3. If the determination in step S4 is NO, the process proceeds to step S5 to determine whether the PLAY button has been input. Where Y
In the case of ES, the process proceeds to step S9, and the reproduction process (PL
AY processing) and returns to step S3.

If the determination in step S5 is NO, the flow advances to step S6 to determine whether the SKIP button has been pressed. If the determination is YES, the flow shifts to step S10 to perform SKIP processing and then to step S3. Return. SKI
P processing is performed for the file No. Is the process of changing the SKI
File No. displayed every time the P button is pressed. Is changed.

If the determination in step S6 is NO, the process proceeds to step S7 to determine whether there has been no operation input for 5 minutes or more. If NO, the process returns to step S3, and if YES, the process returns to step S3. The process proceeds to S20 to set the power saving mode. After that, the power saving mode is maintained while determining whether or not any operation input has been made (step S21). If there is any operation input, the determination in step S21 becomes YES and the process returns to step S1.

On the other hand, if YES in the step S3, the process shifts to a step S11 to display an erase mode (ERASE mode). In this ERASE mode,
As shown in FIG. Is displayed, and “ERASE” is displayed in the display area C. Next, it is determined whether or not the SEL button has been input (step S12). If NO, the process proceeds to step S13, and it is determined whether or not the operation input has been performed for 5 minutes or more. If NO, the process proceeds to step S14. Use to select the file to be deleted. The file is selected by sequentially pressing the SKIP button. File No. Flashes when not selected, and is lit when selected. Next, REC
After confirming whether the selected file should be deleted based on whether the button is pressed (step S15), the selected file is deleted (step S16). During erasing, “ERASE” in the display area C blinks.

If the determination in step S13 is YES, the process shifts to step S20 to set the power saving mode. Thereafter, it is determined whether or not there is any operation input (step S21), and the power saving mode is maintained. .
If there is any operation input, the determination in step S21 becomes YES and the process returns to step S1.

If the determination in step S12 is YES, the process shifts to step S17 to display the PC-LINK mode. FIG. 5C in the PC-LINK mode.
"PC-LINK" indicating the PC-LINK mode is displayed in the display area C as shown in FIG.

Next, the process proceeds to step S18, where it is determined whether or not the SEL button has been pressed.
2 and in the case of YES, the process proceeds to step S19 and the PC
A LINK process is performed, and the process returns to step S17.

As can be seen from the above description, in this embodiment, when the SEL button is input twice, the PC-LIN
A K-mode display is performed, and a PC-LINK process is performed when the SEL button is pressed.

FIG. 4 shows one example of the PC-LINK mode.
9 is a flowchart showing details when a feature of a user's voice is registered in advance in one enroll mode. When the enroll mode is set, “ENROLL” is displayed in the display area C as shown in FIG. In step S50, the connection between the PC and the recorder is confirmed, and it is determined whether or not the connection is normally made (step S51). In the case of NO here, the process stands by until the connection becomes normal. When the connection is confirmed to be normal, data communication between the digital recorder and the PC becomes possible. In this case, the determination in step S51 is YES
Then, the flow advances to the next step S52 to display a text for enrollment on the PC. Here, as shown in FIG. 6, a message "Please read this sentence." Is displayed as the first sentence of the 300 sentences prepared, and "Start" for starting the enrollment process is performed. A "button" and a "stop" button for stopping the enrollment process are displayed.

Next, it is determined whether or not an input start instruction has been received from the operator (step S53). Here, RE
Judgment is made based on whether the C button is input. When the REC button is input, the control signal indicating the start of the input
Sent to. Next, in step S54, the operator sets P
By reading out the text displayed on the screen of C, the input of voice through the microphone 1 is started, and the input voice data is transmitted to the PC through the audio output terminal 18. At the same time, the counter is operated to notify the operator that voice input is being performed, and the count value of the display area B of the display unit 14 is incremented (step S55). Instead of operating the counter, display area C
"ENROLL" may be blinked. Next, it is determined whether or not an error has occurred during the input of the voice (step S56). Here, if a sound such as background noise that is far from the operator's voice is input during voice input, it is processed as an error. Also, when the operator presses the STOP button to interrupt the voice input, it is processed as an input error. In the case of an input error, the counting operation of the counter is stopped (step S57), an error display as shown in FIG. 5E is made on the display unit 14 (step S58), and the process returns to step S52 to read out the same sentence. Then, voice input is performed again.

If the determination in step S56 is NO, it is determined in next step S59 whether the voice input of one sentence has been completed. Here, in the case of NO, the process returns to step S55. While there is no input error and the voice input of one sentence is being performed, the counting operation is continuously performed.
When the voice input of one sentence is completed, step S59
Is YES, and then the counting operation of the counter is stopped in step S60.

Next, it is determined whether or not the enroll operation is completed (step S61). If the enroll operation is completed, the process returns. If there is still a document to be enrolled, the next document is displayed on the PC screen (step S61). After S62), the process returns to step S54 to continue the voice input. When a predetermined number of sentences (here, for example, 300) have been read aloud in this way, the determination in step S61 becomes YES, and the process exits the enrollment mode.

According to the above-described embodiment, in the enroll mode, "ENROLL" is displayed as shown in FIG. 5D, and then the counter is operated to increment the count value. The operator can easily recognize that voice input is being performed in the enroll mode. In addition, the counter is normally provided in the digital recorder as an indicator of the progress of recording or playback, but in the present embodiment, this is also used as a sign indicating that voice input is being performed. There is no need to provide a dedicated display space for that.

Further, the enrollment process is performed using operation buttons (here, REC button, STOP button) provided on the operation unit of the digital recorder as a voice input device capable of recording independently. The enrollment process can be performed without clicking the start button or the stop button displayed in (2) with a mouse or the like. Also, an operation button such as a REC button or a STOP button used for setting a mode when the digital recorder is used alone is also used as an operation button for voice recognition, thereby newly providing an operation button dedicated to voice recognition. There is no need to provide them.

The specific embodiments described above include inventions having the following configurations. (1) An audio input device that includes a display unit, is capable of recording independently, and can perform data communication with an external device, receives audio data transmitted from the audio input device, and performs a voice recognition process on the received audio data. And a voice recognition processing device, wherein the voice recognition processing device has an enroll function for registering data for improving the accuracy of voice recognition processing, and a voice is input to the voice input device. While the input voice information is transmitted to the voice recognition processing device and the enrollment operation is being performed in the voice recognition processing device, the display means of the voice input device indicates that the enrollment operation is being performed. 2. The voice recognition system according to claim 1, wherein a predetermined display is performed.

(2) The speech recognition system according to (1), wherein the display indicating that the enrolling operation is being performed is performed by operating a counter.

According to the invention described in the above (1), it is possible to easily recognize the enroll mode.

According to the invention described in (2),
In addition to the effect of the invention described in (1), there is an effect that it is not necessary to provide a dedicated space for displaying that voice input is being performed.

[0039]

According to the first or second aspect of the present invention, it is possible to easily recognize the enroll mode.

According to the third aspect of the present invention, in addition to the effect of the first aspect, it is not necessary to provide a dedicated space for displaying that voice input is being performed. To play.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a digital recorder to which a voice input device according to the present invention is applied.

FIG. 2 is a diagram illustrating an appearance of a digital recorder.

FIG. 3 is a flowchart illustrating details of processing by a CPU based on an operation input from an operator.

FIG. 4 is a flowchart showing details when a user's voice is registered in advance in an enroll mode as one of the PC-LINK modes.

FIG. 5 is a diagram showing a display on a display unit in each mode.

FIG. 6 is a diagram showing a screen display of a PC at the time of enrollment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Microphone, 2 ... Amplifier (AMP), 3 ... Low-pass filter (LPF), 4 ... Analog-digital converter (A / D), 5 ... DSP, 6 ... Control part (CPU), 7 ... Recording medium, 8 ... Operation Input unit, 9: Digital-to-analog converter (D / A), 10: Low-pass filter (LPF), 11: Amplifier (AMP), 12: Speaker, 13: Drive circuit, 14: Display unit, 15: Power supply control unit, 16 ... PC connection terminal 17 ... Amplifier (AMP) 18 ... Audio output terminal

Claims (3)

[Claims] An audio input device capable of recording independently and capable of data communication with an external device, comprising: a display means; and an external device for registering data for improving accuracy of voice recognition processing. Control means for causing the display means to perform a predetermined display indicating that an enrolling operation is being performed when a sound is input, the sound input apparatus comprising: 2. A voice input device comprising a display means, capable of recording independently and capable of data communication with an external device, receiving voice data transmitted from the voice input device, and performing voice recognition on the received voice data. A voice recognition processing device to be processed, and a voice recognition system comprising: the voice recognition processing device has an enroll function for registering data for improving the accuracy of voice recognition processing;
A voice is input to the voice input device, and the input voice information is transmitted to the voice recognition processing device, and while the enrollment operation is being performed in the voice recognition processing device, a display unit of the voice input device is displayed. 2. The voice recognition system according to claim 1, wherein a predetermined display indicating that an enroll operation is being performed is performed. 3. The voice input device according to claim 1, wherein the display indicating that the enrolling operation is being performed is performed by operating a counter.