JPH02244099A - Voice signal processor - Google Patents

Abstract

PURPOSE:To extract voice command words at high S/N from the noises generated by an acoustic appliance by providing a noise cancelation unit and subtracting the output signal of a driving unit which drives a speaker from the electric signal converted by a microphone. CONSTITUTION:The noise cancelation means 22 subtracts the output signal of the driving unit 21 which drives the speaker 11 from the electric signal converted by the microphone 12. Even if, therefore, the noises radiated from the speaker 11 are mingled in the electric signal converted by the microphone 12, the noises are offset and erased by the output signal of the driving unit 21. The voice command words are extracted at the high S/N in this way from the noises generated by the acoustic appliance without degrading the functions of the acoustic appliance.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Field of Industrial Application) The present invention relates to a voice signal processing device that can extract voice command words with a high S/N ratio even in a noisy environment.

(Prior Art) As a recent trend, various devices such as radios, car stereos, air conditioners, etc. have been mounted on moving objects such as passenger cars. Conventionally, these devices have been operated by switching switches, but recently attempts have been made to operate the devices by voice commands.

By the way, these devices can generate noise in the same frequency band as the audio frequency. In particular, audio equipment such as radios and car stereos can generate quite large levels of noise in the audio frequency band. Therefore, when attempting to operate a device using voice commands, the noise generated by the audio device overlaps with the voice command word, causing a problem in that accurate voice recognition cannot be performed.

In order to solve these problems, for example,
Publication No. 29755 etc. states that when performing voice recognition,
A device has been proposed that causes all devices that adversely affect speech recognition to perform an operation to lower the noise level.

(Problems to be Solved by the Invention) However, in such a conventional device, the functionality of the audio equipment deteriorates, and there is a problem in that people sometimes feel uncomfortable.

The present invention has been made to solve these problems, and is a method for extracting voice command words from the noise generated by audio equipment at a high S/N ratio without degrading the functionality of the audio equipment. This is a common technical issue.

[Structure of the invention]

(Means for solving the problem) The first technical means taken to achieve the above-mentioned technical problem is to provide a noise canceling unit connected to a drive unit that drives a speaker and a microphone,
The output signal of the drive unit is subtracted from the electrical signal converted by the microphone.

In addition, the second technical means taken to achieve the above-mentioned technical problem is to install a mixing unit that mixes electrical signals output from each drive unit into a plurality of drive units that drive a plurality of speakers. A noise canceling unit is further connected to the mixing unit and the microphone, and the output signal of the mixing unit is subtracted from the electrical signal converted by the microphone.

(Operation) According to the first technical means described above, the noise canceling means subtracts the output signal of the drive unit from the electrical signal converted by the microphone.

Therefore, even if noise emitted from the speaker is mixed in the electrical signal converted by the microphone, this noise is canceled out and eliminated by the output signal of the drive unit. As a result, only the voice command word is extracted from the electrical signal converted by the microphone, and the intended technical problem is achieved.

Furthermore, according to the second technical means described above, in addition to the first technical means described above, a mixing unit is provided that mixes the electrical signals output from the plurality of drive units. This mixing unit mixes the output signals of the plurality of drive units and then supplies the mixed signals to the noise canceling means. Therefore, even if noises emitted from a plurality of speakers are mixed in the electrical signal converted by the microphone, these noises are canceled out and eliminated by the output signal of the mixing unit. As a result, only the voice command word is extracted from the electrical signal converted by the microphone, and the intended technical problem is achieved.

(Example) Hereinafter, a preferred first example of the present invention will be described with reference to the accompanying drawings.

FIG. 1 shows an example in which a voice input device 13 is used to operate a computer terminal 14 in a room 10. As shown in FIG.

Microphone 12 is connected to audio input device 13 . Further, the voice input device 13 is connected to a computer terminal 14. Further, a speaker 11 is installed in the room 10. The speaker 11 is connected to broadcast equipment 21, and music, call broadcasts, etc. are played from the speaker 11.

The music played from the speaker 11, the calling method, etc. become noise in the audio frequency band and interfere with the operation of the voice recognition unit 23. In order to eliminate the influence of such noise and perform accurate voice recognition, the voice input device 13 is equipped with a noise canceling unit 22 that is compatible with the present invention.

The electric circuit of the voice input device 13 will be explained below with reference to FIG. The voice input device 13 includes a noise cancellation unit 22, a voice recognition unit 23, and a monitor speaker 27.

The noise canceling unit 22 further includes a first audio input circuit 24, a second audio input circuit 25, and a digital signal processing circuit 26.

The first audio input circuit 24 includes a low-pass filter 241, an amplifier circuit 242 . An A/D converter 243 is provided. The microphone 12 is connected to the first audio input circuit 24 . The electrical signal converted by the microphone 12 has high frequency components removed by a low-pass filter 241, and then is amplified by an amplifier circuit 242. The electrical signal amplified by the amplifier circuit 242 is quantized by the A/D converter 243 and converted into a binary code.

In exactly the same way, the second audio input circuit 25 includes low-pass filters 251 . Amplification circuit 252. An A/D converter 253 is provided. The second audio input circuit 25 is connected to the electric wire 15 that connects the speaker 11 and the broadcasting equipment 21 . The output signal of the broadcasting equipment@21 is amplified by an amplifier circuit 252 after high frequency components are removed by a low-pass filter 251.

The electrical signal amplified by the amplifier circuit 252 is converted into an A/D
The converter 253 quantizes and converts it into a binary code.

By the way, the A/D converters 243 and 253 may generate electrical noise called "folding waves" in the process of quantizing analog signals. A/D converter 243,
If 253 generates a folded wave j, the digital signal processing circuit 26 described later will not operate correctly. Occurs when there is. Therefore, in the device of the first embodiment, low-pass filters 241.25i are used to remove electrical signals having high frequencies to prevent "folded waves" from being generated from the A/D converters 243 and 253.

The binary codes quantized by the A/D converters 243 and 253 are sent to the digital signal processing circuit 26. The digital signal processing circuit 26 includes a delay device 261, an adaptive filter 262, and an adder 263.

The delay device 261 temporally delays the binary code output from the first audio input circuit 24 and ensures the causality of the adaptive filter 262.

The adaptive filter 262 monitors the output signal of the adder 263 and adjusts the phase, time delay, amplitude direction, magnitude, etc. of the output signal from the second audio input circuit 25. As a result of this adjustment, the chamber 10 from the speaker 11 to the microphone 12
The time delay that occurs each time a sound wave propagates inside the
The influence of reflected waves generated when sound waves are reflected within the room 10 is estimated, and the output signal of the second audio input circuit 25 is corrected to be as similar to the output signal of the first audio input circuit 24 as possible.

Adder 263 subtracts the output signal of adaptive filter 262 from the output signal of delay device 261. When the output signal of the adaptive filter 262 is subtracted from the output signal of the delay device 261 in the adder 263, the music, call broadcast, etc. output from the broadcasting equipment@21 are offset and deleted.

The output signal of adder 263 is applied to speech recognition unit 23 . The voice L'2 Nlj unit 23 is a device that can recognize limited languages of unspecified speakers.

The voice recognition unit 23 has messages such as “ON J1 “OFF”, “
A voice pattern such as "Douzachu" is registered in advance, and the next stage computer terminal 14 is controlled while handshaking with the speaker using the monitor speaker 27.

When a voice command word is applied to the voice recognition unit 23 at the same time as music, call broadcast, etc. generated by the speaker 11,
Recognition of spoken command words may become impossible. However, in the device of the first embodiment, music, call broadcasts, etc. generated by the speaker 11 are transmitted to the digital signal processing circuit 26.
Since only the voice command word is applied to the voice recognition unit 23, the voice command word can be recognized more reliably.

Next, a second preferred embodiment of the present invention will be described.

FIG. 3 shows an example in which the voice human power device 35 is used to operate the navigation device f34 of the passenger car 30. The device of the second embodiment and the device of the first embodiment have substantially the same configuration. Therefore, only the constituent elements that are different from the apparatus of the first embodiment will be explained in detail, and the detailed explanation of the constituent elements that are similar to the apparatus of the first embodiment will be omitted by giving the same reference numerals.

The microphone 33 is fixed near the driver's seat 31 so that the driver's voice can be easily heard. The microphone 33 is
Like the device of the first embodiment, it is connected to the first audio input circuit 24 of the audio input bag W35. The voice input device 35 is housed under the foot of the passenger seat 32 so as not to be seen.

The voice input device 35 is connected to the navigation device 34.

Furthermore, two speakers 37R and 37L are installed within the vehicle compartment 36. The speakers 37R, 37L are connected to the in-vehicle audio equipment 38, and the speakers 37R, 37L are connected to the in-vehicle audio equipment 38.
Music, radio broadcasts, etc. are played from 37L.

Music, radio broadcasting, etc. flowing from the speakers 37R and 37L become noise in the audio frequency band, and the voice recognition unit 2
This prevents the operation of step 3. In order to eliminate the influence of such noise and perform accurate voice recognition, the voice input device 35 is equipped with a noise canceling unit 22 to which the present invention is applied.

The electric circuit of the voice input device 35 will be described below with reference to FIG. The voice input device 35 includes a noise canceling unit 22, a voice recognition unit 23, a monitor speaker 27, and a mixing unit 39. The noise canceling unit 22, the voice recognition unit 23, and the monitor speaker 27 are completely the same as those in the first embodiment, so a detailed explanation will be omitted.

The device of the second embodiment includes a plurality of speakers 37R and 37L. Therefore, noises emitted from the plurality of speakers 37R and 37L are input to the microphone 33 at different levels. For example, in the example shown in FIG. 3, the microphone 33 is located at a position biased to the right side of the vehicle interior 36, so that the electrical signal converted by the microphone 33 includes noise emitted from the speaker 37L. The noise radiated from the speaker 37R is included at a higher level than the noise emitted from the speaker 37R.

In order to adjust such unbalance, the mixing unit 39 further includes a balancer 391 and an adder 392. The balancer 391 and the adder 392 are connected to the right channel and the left channel of the in-vehicle audio device 38, respectively, and adjust the mixing ratio of the volume of the electrical signals input from both channels according to the position of the microphone 33 in the vehicle interior 36. Adjust. The adjusted electrical signals are mixed by an adder 392 and then applied to the second audio input circuit 25 of the noise cancellation unit 22.

The noise canceling unit 22 converts the electric signal converted by the microphone 33 into a mixing unit 39.
Subtract the output signal of . At this time, a plurality of speakers 37R,
Even if the noise emitted from the mixing unit 37L is mixed, these noises are offset and eliminated by the output signal of the mixing unit 39. As a result, only the voice command word is extracted from the electrical signal converted by the microphone 33.

The extracted voice command word is recognized by the voice recognition unit 23, and controls the next-stage navigation unit 34 while handshaking with the speaker using the monitor speaker 27.

A plurality of speakers 37R137L in the voice recognition unit 23
If the voice command word is applied at the same time as music, radio broadcast, etc. that is generated, it may become impossible to recognize the voice command word. However, in the device of the second embodiment, music, radio broadcasts, etc. generated by the speakers 37R and 37L are erased by the digital signal processing circuit 26, and only the voice command word is applied to the voice recognition unit 23, so the voice command word is recognized more reliably.

Furthermore, since the device of the second embodiment is equipped with a balancer 391, the vehicle interior 3 in which the plurality of speakers 37R and 37L are disposed
It becomes possible to attach the microphone 33 to any position within the 6.

In the devices of the first and second embodiments described above, since the microphone 12° 33 is connected only to the first voice input unit 24, the voice command word generated by the speaker is transmitted only to the second voice input unit 24. input, the first voice input unit 2
5 is never entered. , Therefore, the voice command word is not erased or weakened by the signal processing of the digital signal processing circuit 26.

Furthermore, since only one microphone 12, 33 is required, the cost of the audio processing device can be reduced.

〔Effect of the invention〕

According to the invention described in claims 1 and 2, voice command words can be extracted from the noise generated by the audio equipment at a high S/N ratio without degrading the functionality of the audio equipment.

Further, according to the second aspect of the invention, voice command words can be extracted with a high S/N ratio from the noise generated by a plurality of audio devices.

[Brief explanation of drawings]

FIG. 1 is a perspective view depicting a chamber to which a first preferred embodiment of the present invention is applied. FIG. 2 is a block diagram depicting the electrical configuration of a first preferred embodiment of the present invention. FIG. 3 is a perspective view depicting a vehicle compartment to which a second preferred embodiment of the present invention is applied. FIG. 4 is a block diagram depicting the electrical configuration of a second preferred embodiment of the present invention. 10...Room (predetermined space), ■...Speaker, 2...Microphone, 3...Audio input device, 14...Computer terminal, 5...1!
Line, 21...Broadcast equipment (drive unit), 2...Noise cancellation unit (noise cancellation unit), 3...Voice recognition unit, 4...First audio input circuit, 5...Second Audio input circuit, 6... Digital signal processing circuit, 7... Monitor speaker, 3o... Passenger car, 1...
Driver's seat, 32... Passenger seat, 3... Microphone, 4... Navigation device, 35... Voice input device, 6... Vehicle interior (predetermined space), 7R... Speaker power (first speaker), 7L...
Speaker (second speaker), 8... Car audio equipment (first drive unit, second drive unit), 39.
...Mixing unit (mixing unit), 391...Balancer (balancer).

Claims (3)

[Claims] (1) a speaker that radiates sound waves into a predetermined space, a drive unit that drives the speaker, a microphone that converts the sound waves propagating in the predetermined space into an electrical signal, and is connected to the drive unit and the microphone,
An audio signal processing device comprising: a noise canceling unit that subtracts an output signal of the drive unit from an electrical signal converted by the microphone. (2) a first speaker that radiates sound waves into a predetermined space; a second speaker that radiates sound waves into the predetermined space; a first drive unit that drives the first speaker; and a second speaker that radiates sound waves into the predetermined space. a second drive unit that drives the speaker; a microphone that converts sound waves propagating in the predetermined space into electrical signals; and a microphone that is connected to the first and second drive units and drives each of them. a mixing unit that mixes electrical signals output from the unit; and a noise canceling unit that is connected to the mixing unit and the microphone and that subtracts the output signal of the mixing unit from the electrical signal converted by the microphone. An audio signal processing device comprising: (3) The mixing unit further includes the first,
The audio signal processing device according to claim 2, further comprising a balancer that changes the mixing ratio of the electric signals output from the second drive unit.