JPH0723400U - Audio equipment - Google Patents

Abstract

(57) [Abstract] [Purpose] To ensure voice control of audio equipment. [Structure] The voice from the microphone is recognized and executed by the voice recognition unit 7. On the other hand, the signal to the speaker is also recognized by the same recognition means 6. When the recognition means 6 recognizes, the prohibition circuit 8 prohibits the output of the recognition means 7. Further, the signal of the microphone is canceled through a circuit having a characteristic almost the same as the transmission characteristic from the speaker 2 to the microphone 3. As a result, a signal having the same sound as the command in the reproduced sound does not malfunction. It also prevents a situation in which voice recognition cannot be performed due to the loud sound of the speaker entering the microphone.

Description

[Detailed description of the device]

[0001]

[Industrial applications]

 The present invention relates to improvements in devices that include voice-controlled audio playback.

[0002]

[Prior art]

 In order to control the audio device when it is difficult to operate manually, there is a means to control by voice input by voice recognition. In this case, a specific voice, for example, a word sound such as “play” or “stop” is registered, and the voice corresponding to the voice is input to operate.

[0003]

[Problems to be solved by the device]

 However, in this case, when a similar word sound was reproduced by the speaker during music reproduction or radio broadcasting, it could be indistinguishable from the operator's command and operated through the microphone. Also, in the case of playing with a loud sound, there was a case where a reproduced sound more than the command voice was input to the microphone and the command sound could not recognize the reproduced sound and the mixed command sound.

[0004]

[Means for Solving the Problems]

 The present invention is intended to obtain a device that can reproduce the same voice as the registered voice during playback, or can correctly separate the command voice even during playback with a loud voice.The voice recognized from the signal through the microphone. Prepare a recognition means and an equivalent recognition means. By using this recognition means, the recognition output is obtained from the signal emitted by the speaker, so that the former does not recognize when the latter recognizes. This can prevent malfunction. The other is to provide a means for canceling the voice signal emitted from the speaker and the voice signal from the speaker input from the microphone, and input only the command sound input to the microphone to the voice recognition means. To prevent interference due to sound emission.

[0005]

[Action]

 As a result, even if the same sound as the command is reproduced in the reproduced sound, no malfunction occurs, and the command can be surely recognized and executed even when reproduced with a loud sound.

[0006]

【Example】

 FIG. 1 shows a block diagram of an embodiment of the present invention. The command signal from the microphone 3 is passed through the amplifier 4 and a command code corresponding to the signal input by the voice recognition unit 7 is output. This voice recognition unit 7 is publicly known and omitted, but at present, there is one that can input a specific voice and register it, and output the registered registration number when the same voice input as the registered voice is input. It is generalized in LSI and the like.

Thus, for example, if a voice input signal “stop” is registered as a “stop” command, and a voice signal “stop” comes in, a fetching stop is executed as a stop command.

In the circuit block diagram of FIG. 1, the reproduction signal input IN is amplified by the amplifiers 1 a and 1 b and is emitted by the speaker 2. A voice recognition unit 6 equivalent to the voice recognition unit 7 is provided for this voice signal, and the signal emitted by this is recognized.

When a “stop” voice is recognized from the emitted signal, the prohibition circuit 8 prohibits the output of the voice recognition unit 7. As a result, the "stop" voice is emitted, and even if the voice recognition unit 7 recognizes the stop command via the microphone, it is also recognized by the voice recognition unit 6 and is prohibited and the stop command is not executed. When the command detection output from the reproduced signal is not output from the voice recognition unit 6 but is performed only from the voice recognition unit 7, the input is made only from the microphone, and the prohibition circuit 8 does not inhibit the execution.

As a result, even if the same command word is reproduced from the reproduced sound, no malfunction occurs and the operation can be reliably performed only by the command sound from the microphone.

On the other hand, the reproduction sound signal from the microphone 3 is reduced by the cancellation circuit 5 according to the reproduction signal emitted to the speaker 2 so that the reproduction sound from the speaker 2 is not disturbed. The cancel circuit 5 is provided with a volume 5-1 in order to make the level difference due to the position from the speaker 2 to the microphone 3 the same, and adjusts the signal from the amplifier 4 and the signal level from 1 to the same level. In order to extract only the band required for voice recognition from these signals, the bandpass filters 5-2 and 5-3 are passed through each other to remove the frequency band unnecessary for recognition, thereby improving the S / N ratio.

Further, a variable delay line 5-4 is provided to match the delay of the sound from the speaker 2 to the microphone 3, and the variable delay line 5-4 is subtracted by the differential amplifier 5-5. As a result, the level meter 10 or the like confirms that the output of the differential amplifier 5-5 is minimized by making adjustments according to the distance from the speaker 2 to the microphone 3.

FIG. 2 shows another embodiment of the cancel circuit 5. The digital signal processor 5'can perform the level adjustment by the volume 5-1 shown in FIG. 1, the filters 5-2, 5-3, the delay line 5-4, etc. by digital circuit processing. In addition to this, it is possible to realize characteristics similar to the characteristics from the speaker 2 to the microphone 3 including the transfer characteristics of the room space, and further increase the canceling effect. Also, when the device is set, an oscillator 11 such as an impulse signal or a sweep signal is added to the input of the amplifier 1, for example, X in FIG. 1 and sound is emitted from the speaker, and the DSP automatically outputs by the program of the DSP so that the output is minimized. You can also tune it.

FIG. 3 shows a block diagram of another embodiment of the present invention. A target disc and song number are selectively played back by a playback unit 1'comprising a changer playback device such as an LD or CD, and a signal amplified by an amplifier 4 from a microphone 3 is mixed with the signal by a mixer 1c to be amplified by the reproducing unit 1 '. Sound more.

Here, the signal of this microphone circuit is taken out and added to the voice recognition unit 12. As for the voice of the voice recognition unit 12, if there are 10 tones such as "zero" and "ichi-cue", it can be operated in the same way as the 10 key, and the number of sounds and the time are short, so it is very easy and error You can recognize less. However, if there is a numbered lyrics while singing, a malfunction such as searching for the song number corresponding to the number in the lyrics will occur. Therefore, in the present invention, it is possible to control only the state in which the reproducing means is not reproducing, such as stopping, and prevent these malfunctions.

Further, as shown in the flowchart of FIG. 4, when voice is input and the reproducing unit such as the changer unit is in a stopped state, the reproducing unit outputs a stop signal 12a to enable the voice recognizing unit 12 to perform recognition. .

When the voice signal from the microphone 3 is recognized from the voice signal 12b of the target numeral, the music number command signal 12c corresponding to the number is output, and the music number is selected by the reproducing unit 1'and the reproduction is started. . At this time, if there is a voice input from the microphone 3, the recognizing unit does not operate because the reproducing unit 1'is not in the stop state, and it is determined that the singing is being performed. When singing ends and a number sound is generated, the corresponding song is played. As described above, the playback of the next song number can be controlled by voice even with only a singing microphone, and a simple and malfunction-free device can be realized.

[0018]

[Effect of device]

 As described above, according to the present invention, a large amount of sound from the speaker is input to the microphone, the voice of the command cannot be recognized, and it is possible to execute the command more reliably.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

FIG. 2 shows another embodiment of the cancellation circuit unit 5 of FIG.

FIG. 3 is another embodiment of the present invention

[Fig. 4] Flow chart

[Explanation of symbols]

1 amplifier 2
Speaker 3 Microphone 4
Amplifier 5 Cancel circuit 5-1
Volume 5-2, 5-3 Band pass filter 5-4
Delay line 5-5 Differential amplifier 6,7
Speech recognition section 8 Prohibited circuit 9
Control output 10 Level meter 11
Oscillator

Claims (4)

[Scope of utility model registration request] 1. An audio device for controlling the operation of a device by inputting voice from a microphone,
First voice recognition means for recognizing specific voice information from a voice signal emitted to a speaker, second voice recognition means for recognizing specific voice information from a signal from a microphone, and first and second voice recognition described above. An audio apparatus comprising: means for adjusting the outputs of the means to be a control output, and when there is an output from the first voice recognition means, an output from the second voice recognition means is prohibited. 2. A circuit for adding a signal emitted from a speaker through a microphone by controlling the signal level and phase of the signal emitted to the speaker so that the signal level and the phase of the signal are inverted. 2. The audio device according to claim 1, wherein a voice input from the microphone through the voice recognition means is recognized and used as a control signal. 3. The audio device according to claim 1, wherein the microphone is a vocal microphone for singing a karaoke song, and the karaoke system is controlled by a control output. 4. The audio apparatus according to claim 3, wherein a control output is output by voice input from a microphone only when the reproducing means of the karaoke system is stopped.