JPH0462598A - Voice level display system for voice recognition device - Google Patents

Abstract

PURPOSE:To display a voice level relatively to a noise level that a voice recognition means sets independently by a voice level display means by providing a level adjusting means which adjusts a voice signal level and controlling the level adjusting means by the voice recognizing means. CONSTITUTION:A voice signal which is inputted from a microphone 3 is amplified by a microphone amplifier 5 and sent to a voice level display circuit 2 through a voice recognizing circuit 1, a buffer amplifier 4, and a level adjusting circuit 6. The voice which is inputted to the circuit 1, on the other hand, contains an ambient noise as well as a voice. Consequently, the circuit 1 regards the mean value of the voice level in a constant time as a noise level 7 except during the reception of a voice command. It is therefore decided that 1 is the head of a word (word head detection 9) when the voice level 8 of the input voice command is relatively larger than the level 7. Here, the circuit 1 controls the level adjusting circuit 6 according to the quantity of variation in this level 7 to display how much the voice level is larger than the noise level on the circuit 2.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a method for displaying a speech level in a speech recognition device.

(Prior Art) Conventionally, there are two types of speech recognition devices: a speaker-specific method, which requires the user's voice to be registered before use, and a speaker-independent method, which allows anyone to use the device without registration. In recent years, speaker-specific methods that can recognize several thousand words have been emerging, but in general, testing systems in the FA field, which target several hundred words,
Sorting is often done using systems that manually convey information by voice. Although the speaker-independent method is still at the stage of recognizing a few dozen words, it is being applied as a voice response system using telephones for banking balance inquiries, transfer notifications, etc.

Such a voice recognition device is provided with a voice level display circuit or the like in order to display the voice level of the operator. A partial configuration diagram of this conventional speech recognition device is shown in FIG. Third
In the figure, 1 is a voice recognition circuit, 2 is a voice level display circuit for displaying the operator's voice level, 3 is a microphone for inputting voice commands, 4 is a buffer amplifier, and 5 is a microphone amplifier. Here, the speech recognition circuit 1 is connected to a control circuit (not shown) that performs control necessary for speech recognition processing.

Next, a conventional audio level display method will be explained.

A voice signal input by an operator through a microphone 3 is amplified by a microphone amplifier 5, passes through a voice recognition circuit 1 and a buffer amplifier 4, and is transmitted to a voice level display circuit 2.

Therefore, the display level of the audio level display circuit 2 is the same as the audio level input manually into the microphone 3, and the operator can see this level display and utter the audio command at an appropriate volume. .

(Problems to be Solved by the Invention) However, the conventional audio level display method described above has the following problems. Normally, voice recognition devices constantly monitor input signals (including voice commands as well as ambient noise, circuit noise, etc.) as a means of detecting the beginning of a word, and this average level is regarded as the noise power, and this level is set in advance. The time when the value exceeds the specified value is considered to be the beginning of a word.

Therefore, when a voice command is uttered in a place with low ambient noise, even a low voice will be recognized, but in a place with high ambient noise, the voice command will not be recognized unless the voice command is uttered in a louder voice than the noise. However, the voice level display circuit simply displays the signal level input from the microphone, and has a problem in that it is not very useful for confirming the appropriate voice command utterance level.

SUMMARY OF THE INVENTION An object of the present invention is to solve these conventional problems and provide a voice level display method in a voice recognition device that can display a voice level that allows voice commands to be uttered at an appropriate volume.

(Means for Solving the Problems) In order to achieve the above object, a voice level display method in a voice recognition device of the present invention provides a voice level display method in a voice recognition device having a voice recognition means and a voice level display means for input voice. A level adjustment means is provided for adjusting an audio signal level for display, and the voice recognition means controls the level adjustment means and uses the voice level display means to adjust the noise level independently set by the voice recognition means. It is characterized by the relative display of audio levels.

(Function) In the present invention, the voice recognition means controls the level adjustment means, and the voice level display means displays the voice level relative to the noise level independently set by the voice recognition means. .

(Example) Hereinafter, an example of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a diagram for explaining an audio level display method showing an embodiment of the present invention. This shows a partial configuration of a speech recognition device. Further, FIG. 2 shows the relationship between the sound level and the calculated noise level in the apparatus of FIG. 1. The vertical axis shows the audio level, and the horizontal axis shows time.

In FIG. 1, 1 is a voice recognition circuit, 2 is a voice level display circuit, 3 is a microphone, 4 is a buffer amplifier, 5 is a microphone amplifier, and 6 is a level controlled by the voice recognition circuit 1 which is characteristic of the present invention. This is an adjustment circuit. here,
Except for the level adjustment circuit 6, the configuration is the same as the conventional one, so please refer to the description of the conventional technique for details.

A voice signal input by an operator through a microphone 3 is amplified by a microphone amplifier 5, and transmitted to a voice level display circuit 2 through a voice recognition circuit 1, a buffer amplifier 4, and a level adjustment circuit 6.

On the other hand, as shown in FIG. 2, the voice input to the voice recognition circuit 1 includes ambient noise in addition to the voice.

For this reason, the voice recognition circuit 1 sets the average voice level over a certain period of time to the noise level 7, except when accepting voice commands. Therefore, the voice level 8 of the voice command is
If the input signal is at a relatively high level with respect to the noise level 7, the speech recognition circuit 1 determines that it is the beginning of a word (word beginning detection 9).

Here, the voice recognition circuit 1 controls the level adjustment circuit 6 in accordance with the amount of change in the noise level 7, and displays on the voice level display circuit 2 how loud the voice level is relative to the noise level. As a result, no matter where there is ambient noise, if the operator utters a voice command to reach a certain display state, the voice recognition circuit 1 will receive an appropriate voice level.

(Effects of the Invention) As explained above, according to the present invention, the speech level detected by the speech recognition device at the beginning of a word can be directly checked using the level display circuit, making it suitable for speech recognition regardless of the level of ambient noise. The voice level allows you to easily pronounce voice commands.

[Brief explanation of drawings]

FIG. 1 is a diagram for explaining the voice level display method in a voice recognition device showing an embodiment of the present invention, FIG. 2 is a diagram showing the relationship between the voice level in the voice recognition device and the calculated noise level, and FIG. The figure is a diagram for explaining a speech level display method in a conventional speech recognition device. 1: Voice recognition circuit, 2: Voice level display circuit, 3: Microphone, 4: Buffer amplifier, 5: Microphone amplifier, 6: Level adjustment circuit, 7: Noise level, 8:
Speech level, 9: Word beginning detection. Neiaki Tokezaka 1 Tofu V Episode 1

Claims (1)

[Scope of Claims] A speech recognition device having a speech recognition means and a speech level display means for input speech, further comprising: a level adjustment means for adjusting a speech signal level for displaying the speech level; A voice level display method in a voice recognition device, characterized in that a level adjustment means is controlled, and the voice level display means displays a voice level relative to a noise level independently set by the voice recognition means.