JPS58152298A - Voice input controller - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a voice input control device, and more particularly to an improvement of a voice input control device that controls controlled equipment such as in-vehicle equipment in response to an input @-.

FIG. 1 is a schematic block diagram of a voice input control ML device which is the background of the present invention. First, the conventional voice input control 1lIII snow will be briefly explained with reference to FIG. Sound is input to microphone 1 and converted into an audio signal,
This voice signal is input to the voice recognition circuit 100. In the speech recognition circuit 100, predetermined speech keywords for selecting possibilities of each controlled*m device and acoustic feature parameters corresponding to device names are registered. Then, the voice recognition 1 circuit 100 analyzes and extracts the waveform of the voice input from the microphone 1, and compares the characteristic parameters with the characteristic parameters registered in advance to identify the corresponding keyword. The control circuit 200 generates a control signal corresponding to the keyword as a recognition result of the speech recognition circuit 100, and supplies it to the corresponding drive circuit 300. Accordingly, the drive circuit 300 controls the controlled device 400.

FIG. 2 is a schematic block diagram of the speech recognition circuit 100 shown in FIG. 1. Next, referring to FIG. 2, the speech recognition circuit 10
0 will be explained. A voice signal from the microphone 1 shown in FIG. 1 is given to a voice analysis circuit 101. This speech analysis circuit 101 includes a plurality of bandpass filters,
It consists of a smoothing circuit, an A-D converter, etc., and analyzes the waveform of a human-generated audio signal to extract characteristic parameters. The feature parameters extracted by this speech analysis circuit 101 are stored in the feature parameter register 102. The registered parameter memory 104 is stored in the IIJw device 4.
It stores feature parameters serving as keywords for all sounds needed to control 00. The similarity calculation circuit 103 has the feature parameters of the input voice stored in the feature parameter register 102 and registered in the registered parameter memory 104. This calculates the degree of similarity with any parameter. The determination circuit 105 determines whether the similarity calculated by the similarity calculation circuit 103 is the highest and compares the similarity at that time with a predetermined threshold [1 ( It identifies which registered pattern the voice corresponds to.Then, the determination circuit 105
' encodes the keyword number as the identification result and outputs it as numerical data. Note that the timing signal generation circuit 106 generates a timing signal for determining the procedure of the series of recognition processing described above, and this timing signal is transmitted to the speech analysis circuit 1'01, the registered parameter memory 104, and the determination circuit 105. give to.

and C/3't-1 The speech recognition circuit 'too constructed as described above has the ability to sufficiently recognize input speech when used in a low-noise environment. However, in places where engine noise is loud, such as inside a car, the recognition rate of the input voice decreases and the noise itself is mistakenly used as a keyword sound. Often strict.

For this reason, when such an audio aim circuit 100 is used in an audio input control device, sufficient performance cannot be obtained. Especially in an environment where a radio is playing or people are talking, some of the waveforms of these sound sources may be similar to the sounds of the registered keywords.
The problem is that the probability of being recognized is particularly high.
is.

Therefore, the main object of the present invention is to provide a voice input control device that can reduce erroneous recognition as much as possible even in a relatively noisy environment and prevent malfunctions of the controlled input. It is to be.

To summarize this invention, a predetermined sound for controlling the controlled device-IIIB is emitted twice, and the sound inputted first is emitted twice. ! Identification results and the next input audio!
! III.
It is configured so that m devices can be used as Ill m devices.

The above objects and other objects and features of the present invention will become more apparent from the detailed description given below with reference to the drawings.

FIG. 3 is a schematic block diagram of an embodiment of the present invention.

Generally, when a person pronounces the same word in succession, the waveforms of those voices are extremely similar, and when the person using this device recognizes each voice pronounced with the intention of controlling, the recognition The probability that the results will be inconsistent is extremely small. Focusing on this point, in this embodiment, the above-mentioned first
The voice input control device shown in FIG. 1 and FIG. Then, the recognition result register 110 stores the recognition result of the second voice pronounced by the voice recognition circuit 100, and the recognition old result register 111
stores the recognition result of the voice first pronounced by the voice recognition circuit 100. Then, the comparison circuit 112 checks whether the recognition result of the second voice pronounced as the content of the new recognition result register 110 matches the recognition result of the first voice pronounced stored in the old recognition result register 111. Discern. When the comparison circuit 112 determines that the two match, it activates the control circuit 200.

Next, the operation will be explained. The speaker produces twice the necessary sound to turn on the power of, for example, the radio 401 as the controlled device. When the first voice is produced, the microphone 1 provides a voice signal corresponding to the voice to the voice recognition circuit 100.

As explained in Fig. 2 above, the speech recognition circuit analyzes the waveform of the speech, extracts the feature parameters, and determines the degree of similarity with the feature parameters of multiple types of speech registered in advance. Recognize the Then, the recognition result of the first voice pronounced is stored in the new recognition result register 110. When the speaker continues to pronounce the same sound -, the speech recognition circuit 100 recognizes the sound. At this time, the I! of the first voice stored in the new recognition result register 110!
The result is transferred to the old recognition result register 111. Then, the new recognition result register 110 stores the recognition result of the second voice produced. Comparison circuit 112 recognizes the new result and compares it with the contents of register 110 and IIM old result register 1.
11, and if they do not match, control circuit 2
00 is disabled (hi), and if they match, the control circuit 200 is enabled. Therefore, the control circuit 200 activates a when the results of the I!
A control signal is generated and applied to the drive circuit 300. depending on,
Drive circuit 300 powers on radio 401 .

In addition, in order to control the radio 401, approximately 20 R sounds are required for turning on the power, selecting the frequency, changing the volume, etc., but the recognition result encoded by the reputation recognition circuit 100 is 5 bits. It is information. Therefore, the new recognition result register 1106 and the old recognition result register 111 can be constructed relatively easily by using some 7 lip-flops, and the comparator 112 can be constructed by using some exclusive OR and AND elements. can do.

In this way, in the embodiment shown in FIG. 3, when the same voice is pronounced twice in succession, the radio 401 can be controlled only if the recognition results of the two voices are the same. Particularly in relatively noisy environments such as the inside of a car, the noise level is high and there are various types of noise, but generally the same #! It is rare for the relevant audio waveform to occur twice in succession. Therefore, the voice input control device according to this embodiment can prevent malfunctions caused by noise.

FIG. 4 is a schematic block diagram of another embodiment of the invention. In the embodiment shown in FIG.
Every time you try to control 01, you have to pronounce the same voice twice in a row, which may cause some inconvenience. Further, the time from when the user intends to control the radio 401 to when the radio 401 actually operates becomes slightly longer.

The embodiment shown in FIG. 4 attempts to eliminate this problem.

That is, in this embodiment, sounds as keywords are classified into at least two types of hierarchy. - The one voice is a keyword group required at the start of a series of 61111 operations such as powering on the controlled equipment, and the other voice is a group of keywords required at the start of a series of 61111 operations such as powering on the controlled equipment,
This is a group of keywords that are frequently used in a short period of time, such as for channel selection and volume adjustment. Correspondingly, the speech recognition circuit 100 includes registered parameter memories 104 and 15 that store feature parameters of at least two types of keywords.
4 are provided. Furthermore, a classification circuit 107 is provided for classifying the keyword recognized by the speech recognition circuit 100 into which of the two types of hierarchy.

This classification circuit 107 uses registered parameter memories 104 and 154 output from the timing signal generation circuit 106.
7. It demultiplexes the memory address signals of 1 and can be constructed by combining several logic gate elements. L, and the classification circuit 107 includes the judgment circuit 1.
The recognition result output is given from 05. Classification circuit 107
functions to force the comparison circuit 112 to output a matching signal when the determination circuit 105 determines that the input sound - is the second keyword group.

Conversely, if the input voice is of the second keyword group, the comparison circuit 112 is caused to perform the same operation as in the embodiment shown in FIG. 3 above. For this reason, the comparator circuit 112 can be modified to the extent that, for example, a two-manpower gate element is replaced with a three-manpower gate element.

By configuring as described above, when starting an operation such as turning on the power, the control operation is achieved only after the same voice is pronounced twice, and the voice is pronounced once for the second keyword group. Control operations can be achieved by simply

Here, the operation of this embodiment will be explained using an example of the voice input procedure.

"Nishiki" as a keyword for the call and a keyword for turning on the radio 401 [Radio J is the first keyword group, and a keyword corresponding to operations such as tuning the radio 401, for example "1 channel" is the second keyword group. It is classified into

The user first pronounces the same keyword twice in a row, such as "L-nshiki" and "ninshiki." This keyword is recognized by the speech recognition circuit 100, the first recognition result is stored in the new recognition result register 110, and when the second speech is recognized, the contents of the new recognition result register 1]0 are changed to the old recognition result register 111. 1 of the sounds transferred to and pronounced for the second time! ! Kenji Song recognized new result register 110
is memorized. At this time, the classification circuit 107 classifies the pronounced voice as belonging to the first keyword group, and the comparison circuit 107 classifies the pronounced voice as belonging to the first keyword group.
12 does not force a match signal to be output. However, when the comparison circuit 112 determines that the content of the aS new result register 110 and the content of the recognized old result register 111 match, it provides a match signal to the control circuit 200. depending on,
Control circuit 200 outputs a control signal for operating drive circuit 300. As a result, the drive circuit 300 is activated, and the radio 401 can be controlled from now on.

Next, the user selects the word "radio" from the second keyword group,
When "Radio J" is pronounced, the classification circuit 107 determines that the sound belongs to the first keyword group, and the comparison circuit 112 outputs a matching signal.
00 provides a control signal to the drive circuit 300 for turning on the power of the radio 401. In response, drive circuit 300 powers on radio 401.

Next, the user pronounces a voice for selecting channels such as "1 channel" or "5 channel" only once. The classification circuit 107 classifies the pronounced voice as belonging to the second keyword group, and forces the comparison circuit 112 to output a matching signal. Based on the recognition result stored in the new recognition result register 110, the control circuit 200 generates control and signals for receiving one channel on the radio 401, for example. Accordingly, the drive circuit 301
to receive one channel.

As described above, according to this embodiment, at the start of a series of control operations such as powering on the radio 401, the same voice is emitted twice, and the delay at the start of the control operation is performed using a tag such as a channel selection operation.
For items that require control, it is only necessary to emit the voice once, so the user does not feel inconvenienced, and the time from when the user intends to control until the radio 401 actually performs the desired operation is shortened. The opening time can be shortened.

Note that the audio ix* circuit 100 in the above-mentioned embodiment. Control circuit 200. I! If the identification result register 110.degree. 111, comparator 112, etc. are formed on the same semiconductor chip, the entire device can be made smaller.

As described above, according to the present invention, the voice for controlling the controlled device is emitted at least twice, and the target controlled device is controlled only when the characteristic parameters of each voice match. Therefore, it is possible to prevent the voice V&Ken from malfunctioning due to the noise of the same song. Moreover, it is controlled by the garage input iw+
Since the ima can be controlled, practical functions can be provided without requiring any manual operation.

[Brief explanation of the drawing]

FIG. 1 is a schematic block diagram of an acoustic input system 1 which is the background of the present invention. . FIG. 2 is a schematic block diagram of the speech recognition circuit shown in FIG. 1. FIG. 3 is a schematic block diagram of one embodiment of the present invention. 141V is a schematic block diagram of another embodiment of the present invention. In the figure, 1 is a microphone, 100 is a speech recognition circuit, 102 is a feature parameter register, 104.154 is a registered parameter memory, 107 is a 9-supply circuit, 110 is a recognition result register, and 111 is l! Old results register, 1
12 is a comparison circuit, 200 is a control circuit, 300 is a drive circuit, and 401 is a radio. Agent Shin Kuzuno - (1 other person) Procedural amendment (voluntary) 1 day i'r Ii' l Shinden I ゛1 eye 1 υ table small Tokugansho Is? -858
119 No. 2 Invention of MRl voice input control device,'
51111 +1: 'i-4'; L and ・Ii f'
Relationship with l 11'1 Applicant 6, Contents of amendment

Claims (1)

[Scope of Claims] (1) A voice input control unit that controls a controlled device in response to input of a predetermined voice, which outputs a voice signal in response to input of voice. @multiple input means, which analyzes the waveform of the audio signal from the audio input means to extract feature parameters, determines the degree of similarity with the feature parameters of multiple @im registered in advance, and sends the same to the input means. Speech recognition means for recognizing input speech; first storage means for storing the recognition result of the speech recognition means; 11 for storing the previous recognition result by the speech recognition means;
2 storage means, a comparison means for comparing the current recognition result of the first storage means and the previous recognition result of the second storage means to determine a match, and a recognition result of the voice recognition means and An audio input angle control device, comprising a control means for controlling the ll1II11 equipment according to the matching output of the comparison means. (2> 1. The predetermined sound includes a first type of voice and a second type of voice, and the voice ggm means includes a first type of sound that stores characteristic parameters of the first type of voice.
a second feature parameter storage means for storing the second type of speech feature parameter; and a feature parameter of the input speech is stored in the first or second feature parameter storage means. and a determining means for determining whether the input voice characteristic parameter is similar to any of the voice characteristic parameters of the first IIWA, and the control means determines whether the input voice characteristic parameter is similar to the first IIWA voice characteristic parameter. Claim 11, wherein when the determining means determines that this is the case, the target device is controlled in accordance with the input voice even if the comparing means derives a mismatch output signal. Described iit\i input control device-0