JP5116440B2 - Sound control device - Google Patents

Description

  The present invention relates to a sound control apparatus having a voice recognition system and an ANC (active noise control) system.

  The sound detection means 12 collects the noise inside the deck of the railway vehicle and the control sound from the speaker 11, and the calculation control means 21 obtains the original noise from the sound obtained from the sound detection means 12 and controls the adaptive filter 24. There is known an ANC device that generates a control sound fed back from a speaker and actively reduces noise (see, for example, Patent Document 1).

  Controls the adaptive filter so that the error signals placed in multiple locations in the car detect the combined signal of the unwanted sound and the cancel sound and cancel the unwanted sound using the synthesized signal and the music signal reproduced from the music playback device. A music playback device is known (see, for example, Patent Document 2).

  There is known an in-vehicle navigation device 300 having a voice recognition unit 110 for analyzing an input of a voice signal from a microphone 4 and operating with an uttered voice of an occupant (see, for example, Patent Document 3). Note that the in-vehicle navigation device 300 includes a noise canceling unit for removing a noise component that enters the microphone 4.

Japanese Patent Laying-Open No. 2005-96498 (FIG. 3) Japanese Utility Model Publication No. 2005-96498 (FIG. 3) Japanese Patent Laying-Open No. 2003-344083 (FIG. 2)

  When a voice recognition system and an ANC system are installed together, the ANC system outputs a canceling sound having a phase opposite to that of the noise in order to reduce the noise at the seat ear position. Then, the noise at the seat ear position (near the error microphone) is output. However, in other cases (for example, near the voice recognition microphone), the noise level is increased. That is, there is a problem that the noise level near the microphone for speech recognition increases.

  Accordingly, an object of the present invention is to provide a sound control device that can solve the above-described problems.

  It is another object of the present invention to provide a sound control device that enables coexistence of an ANC system and a voice recognition system.

  The sound control device according to the present invention includes a first sound detection unit, a voice recognition unit for performing voice recognition based on a first detection signal from the first sound detection unit, a second sound detection unit, Based on the first detection signal according to the switching unit that outputs one of the first detection signal from the sound detection unit and the second detection signal from the second sound detection unit, and the signal output from the switching unit A noise control unit that outputs a noise cancellation signal for canceling noise near the first sound detection unit or a noise cancellation signal for canceling noise near the second sound detection unit based on the second detection signal It is characterized by that.

  The sound control device according to the present invention includes a first sound detection unit, a voice recognition unit for performing voice recognition based on the first detection signal from the first sound detection unit, a second sound detection unit, A noise control unit that outputs a noise cancellation signal for canceling noise in the vicinity of the second sound detection unit based on two detection signals; and a blocking unit that blocks the second detection signal or the noise cancellation signal. To do.

  Furthermore, the sound control device according to the present invention includes a first sound detection unit, a voice recognition unit for performing voice recognition based on the first detection signal from the first sound detection unit, a second sound detection unit, And a noise control unit that outputs a noise cancellation signal for canceling noise in the vicinity of the second sound detection unit based on the two detection signals, and an operation stop unit that stops the operation of the noise control unit.

  According to the sound control device of the present invention, even when the error microphone of the ANC system is different from the microphone for speech recognition, the ANC system can be operated without adversely affecting the speech recognition.

  Further, according to the sound control device of the present invention, the noise level near the voice recognition microphone is reduced, so that the S / N ratio is improved and the voice recognition rate can be increased.

  A sound control apparatus according to the present invention will be described below with reference to the drawings. However, it should be noted that the technical scope of the present invention is not limited to these embodiments, but extends to the invention described in the claims and equivalents thereof.

  FIG. 1 is a schematic configuration diagram of a sound control device 1 according to the present invention.

  The sound control device 1 includes a sound source 10, a voice recognition system 20, a voice recognition microphone 21 that outputs a first detection signal, an error microphone 31 that is arranged near an occupant and outputs a second detection signal, a speaker 32, an ANC ( An active noise canceling unit 33, a summing unit 38, an operation unit 40, a switching unit 41, and the like are included. Also, let C be the spatial transfer characteristic in the space from the speaker 32 to the error microphone 31.

  The sound source 10 can be a playback device such as a CD, a DVD, or an MD, a radio, a TV, a navigation system, or the like.

  The speech recognition system 20 is configured to analyze the content of the occupant's utterance based on the first detection signal from the speech recognition microphone 21 and transmit the recognized content to a navigation device (not shown). .

  The ANC unit 33 includes a noise sensor 34, a C circuit 35, an adaptive filter calculation unit (LMS) 36, an adaptive filter 37, and the like disposed in the vicinity of a noise source 50 such as an engine of a vehicle.

  The operation unit 40 includes a switch button or the like disposed in the vicinity of the passenger. When the operation unit 40 is turned on, voice recognition by the voice recognition system 20 is started. When the operation unit 40 is turned on, the first detection signal from the voice recognition microphone 21 is input to the ANC unit 33 by switching the switching circuit 41. The switching circuit 41 can be configured by a semiconductor element, a contact switching type switch, or the like.

  Next, the operation of the ANC unit 33 will be described.

  In the ANC unit 33, when the operation unit 40 is turned off and the switching circuit 41 inputs the second detection signal from the error microphone 31 to the ANC unit 33, the noise x from the noise source 50 in the vicinity of the error microphone 31 ( A noise cancel signal y (n) for canceling n) is generated.

  When the second detection signal detected by the error microphone 31 is e (n), the output of the C circuit 35 is r (n), and the coefficient of the adaptive filter 37 is w (n), the following expression is established.

  Here, the spatial transfer function in the space from the speaker 32 to the error microphone 31 is C, and the step size parameter of the adaptive filter 37 is μ.

  That is, the adaptive filter calculation unit 36 controls the adaptive filter 37 so as to satisfy the filter update equation shown in Equation (3), thereby canceling noise x (n) from the noise source 50 in the vicinity of the error microphone 31. Therefore, it is possible to output a noise cancel signal y (n) for the purpose.

  The summing unit 38 adds the sound signal a (n) from the sound source 10 and the noise cancellation signal y (n) output from the adaptive filter 37 and outputs the sum to the speaker 32. Therefore, since the noise from the noise source 50 is canceled by the noise cancellation signal y (n), the occupant can view only the sound based on the sound signal a (n) from the sound source 10.

  In this way, the sound control device 1 can realize a good acoustic environment that can cancel the leakage sound from the noise source 50.

  When the operation unit 40 is turned ON and the switching circuit 41 inputs the first detection signal from the speech recognition system microphone 21 to the ANC unit 33, the ANC unit 33 is a noise source in the vicinity of the speech recognition system microphone 21. A noise cancel signal y ′ (n) for canceling the noise x (n) from 50 is generated.

  When the first detection signal detected by the speech recognition system microphone 21 is e ′ (n), the output of the C circuit 35 is r (n), and the coefficient of the adaptive filter 37 is w (n), the following equation is established. .

  Here, it is assumed that the spatial transfer function in the space between the speaker 32 and the speech recognition system microphone 21 is also C, and the step size parameter of the adaptive filter 37 is μ.

  That is, in order to cancel the noise from the noise source 50 in the vicinity of the microphone 21 for the speech recognition system by controlling the adaptive filter 37 by the adaptive filter calculation unit 36 so as to satisfy the filter update equation shown in Equation (6). Noise cancellation signal y ′ (n) can be output.

  The summing unit 38 adds the sound signal a (n) from the sound source 10 and the noise cancellation signal y ′ (n) output from the adaptive filter 37 and outputs the sum to the speaker 32. Therefore, the noise from the noise source 50 is canceled in the vicinity of the speech recognition system microphone 21 by the noise cancellation signal y ′ (n), and the noise level in the vicinity of the speech recognition microphone 21 is reduced. Thus, the speech recognition rate in the speech recognition system 20 can be increased. That is, the voice recognition system 20 can be operated without the ANC unit 33 adversely affecting the voice recognition system 20.

  FIG. 2 is a schematic configuration diagram of another sound control apparatus 100 according to the present invention.

  The difference between the sound control device 100 shown in FIG. 2 and the sound control device 1 shown in FIG. 1 is that the sound control device 100 has a cutoff circuit 101 and is output to the summing unit 38 according to the operation of the operation unit 40. The output of the noise cancellation signal y (n) is cut off. In the sound control device 100 shown in FIG. 2, the same components as those of the sound control device 1 shown in FIG.

  In the sound control device 100 shown in FIG. 2, when the operation unit 40 is turned on, the noise cancellation signal y (n) output to the summing unit 38 is blocked by the blocking circuit 101, so that the ANC unit 33 is substantially set. Is not operating. That is, the ANC unit 33 does not adversely affect the voice recognition in the voice recognition system 20.

  Further, in the sound control device 100 shown in FIG. 2, when the operation unit 40 is turned off and the cutoff circuit 101 does not cut off the noise cancellation signal y (n) output to the summing unit 38, the sound control device shown in FIG. Similarly to 1, the sound control apparatus 100 can realize a good acoustic environment in which the leakage sound from the noise source 50 can be canceled.

  In the sound control apparatus 100 shown in FIG. 2, the cutoff circuit 101 blocks the noise cancellation signal y (n) output to the summation unit 38, but the cutoff circuit 101 responds to the operation of the operation unit 40 and an error microphone. You may comprise so that the 2nd detection signal from 31 may be interrupted | blocked. Even if the second detection signal from the error microphone 31 is cut off, the ANC unit 33 is not substantially operated, and the ANC unit 33 does not adversely affect voice recognition in the voice recognition system 20.

  FIG. 3 is a schematic configuration diagram of still another sound control apparatus 110 according to the present invention.

  The difference between the sound control device 110 shown in FIG. 3 and the sound control device 1 shown in FIG. 1 is that the sound control device 110 is configured such that the operation of the ANC unit 33 is stopped according to the operation of the operation unit 40. It is a point. In the sound control device 110 shown in FIG. 3, the same components as those of the sound control device 1 shown in FIG.

  In the sound control device 110 shown in FIG. 3, when the operation unit 40 is turned on, a stop signal is transmitted from the operation unit 40 to the adaptive filter calculation unit 36, and the operation of the ANC unit 33 is stopped. Therefore, the noise cancellation signal y (n) is not output to the summation unit 38, and the ANC unit 33 does not adversely affect the speech recognition in the speech recognition system 20.

  In the sound control device 110 shown in FIG. 3, when the operation unit 40 is turned off and no stop signal is transmitted to the adaptive filter calculation unit 36, the sound control device 110 is similar to the sound control device 1 shown in FIG. 1. It is possible to realize a favorable acoustic environment that can cancel the leakage sound from the noise source 50.

  FIG. 4 is a schematic configuration diagram of still another sound control device 120 according to the present invention.

  The difference between the sound control device 120 shown in FIG. 4 and the sound control device 1 shown in FIG. 1 is that, in the sound control device 120, the switching unit 121 switches the C / C ′ circuit 122 according to the operation of the operation unit 40. It is only the point that is configured to do. In the sound control device 120 shown in FIG. 4, the same components as those of the sound control device 1 shown in FIG.

  In the sound control apparatus 1 described above, the spatial transfer function in the space from the speaker 32 to the speech recognition microphone 21 and the spatial transfer function in the space from the speaker 32 to the error microphone 31 are considered to be substantially the same. Is going. However, in the voice control device 120, the spatial transfer function in the space from the speaker 32 to the voice recognition microphone 21 is C ′, and the spatial transfer function in the space from the speaker 32 to the error microphone 31 is C. Accordingly, the switching unit 121 is configured to switch the C / C ′ circuit 122. Note that the values of the spatial transfer function C ′ and the spatial transfer function C are specified in advance by measurement or the like and set in the C / C ′ circuit 122.

  When the operation unit 40 is turned off and the ANC unit is operated, the switching unit 121 causes the C / C ′ circuit 122 to set a value corresponding to the spatial transfer function C. When the operation unit 40 is turned off and the ANC unit is operated, the switching circuit 41 operates so as to input the second detection signal from the error microphone 31 to the ANC unit. The second detection signal detected by the error microphone 31 at this time is e (n), the output corresponding to the spatial transfer function C of the C / C ′ circuit 122 is r (n), and the coefficient of the applied filter 37 is w (n). ), Where y (n) is a noise cancellation signal for canceling the noise x (n) from the noise source 50 in the vicinity of the error microphone 31, their relationship is expressed by the aforementioned equations (1) to (3). .

  That is, the noise cancellation for canceling the noise from the noise source 50 in the vicinity of the error microphone 31 by controlling the adaptive filter 37 by the adaptive filter calculation unit 36 so as to satisfy the filter update expression shown in Expression (3). It is possible to output the signal y (n).

  The summing unit 38 adds the sound signal a (n) from the sound source 10 and the noise cancellation signal y (n) output from the adaptive filter 37 and outputs the sum to the speaker 32. Therefore, since the noise from the noise source 50 is canceled by the noise cancellation signal y (n), the occupant can view only the sound based on the sound signal a (n) from the sound source 10.

  In this way, the sound control device 120 can realize a good acoustic environment that can cancel the leakage sound from the noise source 50.

  On the other hand, when voice recognition is started by turning on the operation unit 40, the C / C ′ circuit 122 sets a value corresponding to the spatial transfer function C ′ by the switching unit 121. When voice recognition is started by turning on the operation unit 40, the switching circuit 41 operates to input the first detection signal from the voice recognition system microphone 21 to the ANC unit. The first detection signal detected by the voice recognition system microphone 21 at this time is e ′ (n), the output corresponding to the spatial transfer function C of the C / C ′ circuit 122 is r ′ (n), and the applied filter 37 When the coefficient is w (n) and the noise cancellation signal for canceling the noise x (n) from the noise source 50 in the vicinity of the microphone 21 for the speech recognition system is y ′ (n), the following expression is established.

  That is, in order to cancel the noise from the noise source 50 in the vicinity of the speech recognition system microphone 21 by controlling the adaptive filter 37 by the adaptive filter calculation unit 36 so as to satisfy the filter update equation shown in Equation (9). Noise cancellation signal y ′ (n) can be output.

  The summing unit 38 adds the sound signal a (n) from the sound source 10 and the noise cancellation signal y ′ (n) output from the adaptive filter 37 and outputs the sum to the speaker 32. Therefore, the noise from the noise source 50 is canceled in the vicinity of the speech recognition system microphone 21 by the noise cancellation signal y ′ (n), and the noise level in the vicinity of the speech recognition microphone 21 is reduced. Thus, the speech recognition rate in the speech recognition system 20 can be increased. That is, the voice recognition system 20 can be operated without the ANC unit 33 adversely affecting the voice recognition system 20. In particular, the sound control device 120 performs control by distinguishing the spatial transfer function C in the space from the speaker 32 to the error microphone 31 and the spatial transfer function C ′ in the space from the speaker 32 to the speech recognition system microphone 21. Therefore, more accurate control can be performed.

  In the sound control devices 1, 100, 110, and 120 described above, the operation of the sound control device is switched by the operation of the operation unit 40 so that the ANC unit 33 does not adversely affect the speech recognition in the speech recognition system 20. The operation of the sound control device may be switched by other methods. For example, when a wind pressure sensor is arranged in the vicinity of the voice recognition microphone 21 and the wind pressure sensor detects an output equal to or higher than a threshold value, it is determined that the occupant is speaking to the voice recognition microphone 21. You may comprise so that operation | movement switching of a sound control apparatus may be performed so that the ANC part 33 may not have a bad influence on the speech recognition in the speech recognition system 20. FIG. In addition, a detection unit that detects the output of the first detection signal from the voice recognition microphone 21 is provided, and when the detection unit detects an output that is equal to or greater than a threshold value, the occupant speaks to the voice recognition microphone 21. It may be determined that the operation of the sound control device is switched so that the ANC unit 33 does not adversely affect the speech recognition in the speech recognition system 20.

  In addition, the configuration of the ANC unit 33 in the sound control devices 1, 100, 110, and 120 described above is an example, and other ANC configurations are used to reduce noise near the error microphone when speech recognition is not performed. You may comprise as follows.

1 is a schematic configuration diagram of a sound control device 1 according to the present invention. 1 is a schematic configuration diagram of a sound control device 100 according to the present invention. It is a schematic block diagram of the sound control apparatus 110 which concerns on this invention. It is a schematic block diagram of the sound control apparatus 120 which concerns on this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1,100,110,120 Sound control apparatus 10 Sound source 20 Voice recognition system 21 Voice recognition microphone 31 Error microphone 32 Speaker 33 ANC part 40 Operation part 41 Switching circuit 50 Noise source 101 Blocking circuit 121 Switching part

Claims (3)

A voice recognition unit for performing voice recognition based on a first detection signal from the first sound detection unit and the first sound detection unit;
A second sound detector;
A switching unit that outputs one of the first detection signal from the first sound detection unit and the second detection signal from the second sound detection unit;
In response to a signal output from the switching unit, a noise cancellation signal for canceling noise in the vicinity of the first sound detection unit based on the first detection signal, or the second based on the second detection signal. A noise control unit that outputs a noise cancellation signal for canceling noise in the vicinity of the sound detection unit;
Sound source,
An output unit for outputting the signal from the sound source and the noise cancellation signal;
Switching between the first transfer characteristic between the first sound detection unit and the output unit and the second transfer characteristic between the second sound detection unit and the output unit, which are used in the noise control unit. A transfer characteristic switching unit for performing
A sound control apparatus comprising: A switching signal generator for generating a switching signal;
The sound control device according to claim 1, wherein the switching unit performs output switching between the first detection signal and the second detection signal in accordance with the switching signal.   The sound control device according to claim 2, wherein the switching signal generation unit is an operation switch, a wind pressure sensor, or an output detection unit of the first detection signal.

Images