Patent Grant
7467084
The present invention relates to a method and to a device for operating voice-enhancement systems, such as communication and/or intercom/two-way intercom or duplex telephony devices in motor vehicles, in which voice signals are picked up via a microphone system and routed to at least one loudspeaker.
Methods of, this kind are used in motor vehicles for voice-enhancement duplex telephony or for supporting voice input-controlled electronic or electrical components. The fundamental difficulty that arises is that, depending on the particular operating state, background noise is present in a motor vehicle. It masks the voice commands. Intercom and two-way intercom or duplex telephony systems in motor vehicles are mainly used in large vehicles, minibuses, etc. However, they may also be used in normal passenger cars. When using voice-controlled input units for electrical components in motor vehicles, it is may be important for the background noise to be suppressed or for the voice command to be filtered out.
A voice-recognition device for a motor vehicle is described in European Published Patent Application No. 0 078 014, in which the status of engine operation and/or motor vehicle movement is signaled or fed in, via sensors, to the amplifier system of the voice-recognition device. Based on this, a noise-level control is used to attempt to filter out the voice command from the background noise.
A filtering operation is described in International Published Patent Application No. WO 97/34290, where periodic interfering noise signals are filtered out by determining their periods and by using a generator to interfere with them, so that the voice signal remains.
In German Published Patent Application No. 197 05 471, it is described to support a voice recognition with the aid of transversal filtering.
In German Published Patent Application No. 41 06 405, a method is described for subtracting noise from the voice signal, a multiplicity of microphones being used. A duplex telephony device having a plurality of microphones is described in German Published Patent Application No. 199 58 836.
In German Published Patent Application No. 39 25 589, it is described to use a multiple microphone system, in which, in motor vehicle applications, one of the microphones is placed in the engine compartment and one other microphone in the passenger compartment. A subtraction of both signals then follows. A disadvantage in this context is that only the engine noise or the actual running noise of the vehicle itself is subtracted from the total signal in the passenger compartment. Specific secondary noises are disregarded in this case. Also lacking is a feedback suppression. Everywhere that microphones and loudspeakers are placed in acoustically coupleable vicinity, the acoustic signal that is extracted, coupled out or decoupled at the loudspeaker is fed back, in turn, into the microphone. The result is a so-called feedback, and a subsequent overmodulation. Methods for avoiding such an overmodulation are described in European Published Patent Application No. 1 077 013, International Published Patent Application No. WO 02/069487 and International Published Patent Application No. WO 02/21817.
It is an object of the present invention to provide a method and a device that may improve the verbal communication among the occupants of a vehicle.
The above and other beneficial objects of the present invention may be achieved by providing a method and a device as described herein.
In this context, to operate a voice-enhancement system, such as a communication and/or intercom/two-way intercom or duplex telephony device in a motor vehicle, using at least one microphone and at least one loudspeaker to reproduce a signal generated by the microphone, as well as a bandpass filter configured between the microphone and the loudspeaker, the bandpass filter is adjusted as a function of a comparison between the power of the signal generated by the microphone at a test frequency, and the power of the signal generated by the microphone at an at least substantially integral multiple, thus at essentially a harmonic of the test frequency, or as a function of a comparison between the power of the signal generated by the microphone at a test frequency, and the power of the signal generated by the microphone at a test frequency at at least an earlier point in time. One or more frequencies of the signal generated by the microphone may be suitable as a test frequency. In an example embodiment of the present invention, the frequency at which the power of the signal generated by the microphone is mainly at its maximum, is selected as a test frequency. Alternatively, a plurality of frequency components having substantial power are selected as test frequencies.
In another example embodiment of the present invention, the bandpass filter is adjusted both as a function of a comparison between the power of the signal generated by the microphone at the test frequency, and the power of the signal generated by the microphone at an at least substantially integral multiple of the test frequency, as well as as a function of a comparison between the power of the signal generated by the microphone at the test frequency, and the power of the signal generated by the microphone at the test frequency at at least an earlier point in time.
In another example embodiment of the present invention, the bandpass filter is set to block the component of the signal generated by the microphone, using a stop frequency, (only) when the power of the signal generated by the microphone at the test frequency is greater by more than an upper limiting value than the power of the signal generated by the microphone at the first harmonic of the test frequency. Stop frequency in the context of the present invention may also be a frequency range and not just a single frequency.
In another example embodiment of the present invention, the upper limiting value is between 20 and 40 dB. The upper limiting value may amount to, e.g., approximately 30 dB.
In yet another example embodiment of the present invention, the bandpass filter is set so as not to block the component of the signal generated by the microphone, using the stop frequency, when the power of the signal generated by the microphone at the test frequency is greater by less than a lower limiting value than the power of the signal generated by the microphone at the first harmonic of the test frequency.
In another example embodiment of the present invention, the lower limiting value may be between 5 and 20 dB. The lower limiting value may amount to, e.g., approximately 12 dB.
In another example embodiment of the present invention, by comparing the power of the signal generated by the microphone at the test frequency with the power of the signal generated by the microphone at the test frequency at at least earlier points in time, it may be decided whether the power of the signal generated by the microphone at the test frequency is increasing exponentially.
In yet another example embodiment of the present invention, the bandpass filter is set to block the component of the signal generated by the microphone, at the stop frequency, when the decision is made that the power of the signal generated by the microphone at the test frequency is increasing exponentially.
In another example embodiment of the present invention, the bandpass filter is set to block the component of the signal generated by the microphone, using a stop frequency, (only) when the power of the signal generated by the microphone at the test frequency is greater than a response threshold for longer than a first response time, the first response time, e.g., being greater than, e.g., approximately 750 ms.
In yet another example embodiment of the present invention, the power is determined at more than one test frequency, and the bandpass filter is set to block the component of the signal generated by the microphone, using the stop frequency, only when the power of the signal generated by the microphone at a test frequency is greater than the power of the signal generated by the microphone for longer than a second response time, at every other test frequency, the second response time advantageously being greater than, e.g., approximately 750 ms.
In another example embodiment of the present invention, the adjustment or setting of the bandpass filter with respect to the test frequency is repeated, at the earliest, following a minimum response or dead time. The minimum response time may be, e.g., 200 ms to 300 ms.
In yet another example embodiment of the present invention, the bandpass filter is set to block the component of the signal generated by the microphone at a frequency range around the stop frequency when, following a repetition time, which is greater than the minimum response time, the power of the signal generated by the microphone at the test frequency is greater by more than an upper limiting value than the power of the signal generated by the microphone at the essentially first harmonic of the test frequency, and/or when the decision is made that the power of the signal generated by the microphone at the test frequency is increasing exponentially.
In yet another example embodiment of the present invention, the bandpass filter is set to block the component of the signal generated by the microphone at an expanded frequency range around the test frequency when, following a repetition time, which is greater than the minimum response time, the power of the signal generated by the microphone at the test frequency is greater by more than an upper limiting value than the power of the signal generated by the microphone at the essentially first harmonic of the test frequency, and/or when the decision is made that the power of the signal generated by the microphone at the test frequency is increasing exponentially.
In addition to the foregoing, to operate a voice-enhancement system, such as a communication and/or intercom/two-way intercom or duplex telephony device in a motor vehicle, using at least one microphone and at least one loudspeaker to reproduce a signal generated by the microphone, as well as a bandpass filter configured between the microphone and the loudspeaker, the power of the signal generated by the microphone is defined at at least three test frequencies, it being ascertained by evaluating the power of the signal generated by the microphone, at the test frequencies, whether feedback exists, and the bandpass filter being set to block a component of the signal generated by the microphone that exists around a stop frequency, when it is established that feedback exists.
Stop frequency in the context of the present invention may be the test frequency at which the power of the signal generated by the microphone is at its maximum. In an example embodiment of the present invention, however, the stop frequency is the test frequency, to which a correction frequency is added and at which the power of the signal generated by the microphone is at its maximum; i.e., a correction frequency is added to the test frequency at which the power of the signal generated by the microphone is at its maximum. This correction frequency may be formed as a function of the power of the signal generated by the microphone at the test frequency at which the power of the signal generated by the microphone is at its maximum, as well as a function of the power of the signal generated by the microphone at at least one test frequency existing, e.g., directly, next to this test frequency.
Thus, the correction frequency may be generated in accordance with:
fkorr=sign*fdist*Pmaxneigh/(Pmax+Pmaxneigh), in which:
This is further described on the basis of the following example:
192 test frequencies f1, f2, . . . f192 are assumed. f1 is equal to 40 Hz. fdist is 40 Hz for all test frequencies. In addition, for the powers of the signal generated by the microphone, it holds for the test frequencies f1, f2, . . . f192:
P(f1, f2, . . . f94)=1
P(f95)=4
P(f96)=16
P(f97)=2
P(f98, f99, . . . f192)=1
It then holds that:
fkorr=(−)*40 Hz*4(16+2)=−8 Hz
The test frequency at which the power of the signal generated by the microphone is at its maximum is, thus, 3840 Hz, and the stop frequency is 3832 Hz.
It may be provided that, at least in certain example embodiments, to generate the correction frequency in accordance with:
fkorr=Δf*(Pneighright−Pneighleft)/(Pmax+|Pneighright−Pneighleft|), wherein:
Using the above numerical example as a basis, it holds, therefore, in this case that:
fkorr=40 Hz*(2−4)/(16+|4−2|)=−4.44 Hz
The test frequency at which the power of the signal generated by the microphone is at its maximum is, thus, 3840 Hz, and the stop frequency 3835.56 Hz.
In another example embodiment of the present invention, the spacings between at least some of the test frequencies, or all of the test frequencies, are equidistant.
In yet another example embodiment of the present invention, the existence of feedback may only be ascertained when the power of the signal generated by the microphone at the test frequency at which the power of the signal generated by the microphone is at a maximum, is greater by more than an upper limiting value than the power of the signal generated by the microphone at the first harmonic of this test frequency, the upper limiting value, e.g., being between 20 and 40 dB, for the most part, at, e.g., 30 dB.
In yet another example embodiment of the present invention, the non-existence of feedback is ascertained when the power of the signal generated by the microphone at the test frequency at which the power of the signal generated by the microphone is at a maximum, is greater by less than a lower limiting value than the power of the signal generated by the microphone at the first harmonic of this test frequency, the lower limiting value, e.g., being between 5 and 20 dB, for the most part, at, e.g., 12 dB.
In another example embodiment of the present invention, the existence of feedback is (only) ascertained when the power of the signal generated by the microphone at the test frequency at which the power of the signal generated by the microphone is at a maximum, is increasing, at least approximately, exponentially.
In another example embodiment of the present invention, the existence of feedback is (only) ascertained when the power of the signal generated by the microphone is greater, at at least one test frequency, than a response threshold for longer than a first response time. The first response time may be greater than, e.g., approximately 750 ms. The response threshold may be selected as a function of the power of signal S, i.e., of the sum of the power of all test frequencies.
In another example embodiment of the present invention, the existence of feedback is (only) ascertained when the power of the signal generated by the microphone is greater for longer than a first response time, at at least one test frequency, than the power of the signal generated by the microphone at every other test frequency. The second response time may be greater than, e.g., approximately 750 ms.
In another example embodiment of the present invention, the adjustment or setting of the bandpass filter is repeated, at the earliest, following a minimum response or dead time, which may be, e.g., between 100 ms and 300 ms.
In yet another example embodiment of the present invention, the power of the signal generated by the microphone is determined at at least 50, e.g., at 150 to 300 test frequencies.
In another example embodiment of the present invention, the bandpass filter is a notch filter or a filter bank or multifilter having at least one notch filter. The filter bank may include 10 notch filters, for example.
In accordance with an example embodiment of the present invention, a method for operating a voice-controlled system, such as a communication and/or an intercommunication device for a motor vehicle, including a microphone, a speaker connected to the microphone and a bandpass filter within a signal path between the microphone and the speaker, the bandpass filter including at least one adjustable parameter includes analyzing the frequency of a signal obtained by the microphone. The method also includes at least one of obtaining a comparison of the power at a certain frequency of the signal and the power of at least one harmonic of the certain frequency and obtaining a comparison of the power at a certain frequency of the signal and the power of the certain frequency at a later instant. The method further includes adjusting the at least one adjustable parameter dependent on the comparison.
Further aspects and details are set forth below in the following description of exemplary embodiments.
In the present exemplary embodiment, loudspeakers 9, 17, 18, 19, 20 output a signal generated by microphone 21. Loudspeakers 7, 17, 18, 19, 20 output a signal generated by microphone 22. Loudspeakers 7, 9, 19, 20 output a signal generated by microphone 23. Loudspeakers 7, 9, 17, 18 output a signal generated by microphone 24. In this manner, the possibility of effective verbal communication in a motor vehicle may be enhanced. In principle, the more strongly a signal is amplified between one of microphones 21, 22, 23, 24 and one of loudspeakers 7, 9, 17, 18, 19, 20, the more effective the communication is. However, the possibility of implementing such an amplification is limited by possible feedback effects caused by sound radiated by a loudspeaker 7, 9, 17, 18, 19, 20, which is received by a microphone 21, 22, 23, 24, and is subsequently amplified and radiated by loudspeaker 7, 9, 17, 18, 19, 20.
To reduce such a feedback, as illustrated in
To amplify signal S and/or signal S′, amplifiers may be provided. However, the amplifier function may also be provided by the bandpass filter.
It may be provided to average the power over time at test frequencies fn, fn+1, fn+2, fn+3, fn+4, fn+5, fn+6, fn+7, fn+8, i.e. to form an average value over time, and to analyze this time average of the power instead of the current or active-power of signal S at test frequencies fn, fn+1, fn+2, fn+3Fn+4, fn+5, fn+6, fn+7, fn+8. To the extent that the power of signal S is mentioned herein, it may also include the average value of the power formed over a certain time period. In addition, the concept of power in accordance with the present invention, may also include amplitude or its time average. Also to be included in accordance with the present invention are other variations of the power, of the amplitude, or of their time averages, such as normalized quantities, etc. Thus, for instance, in the context of the present invention, the power of signal S at a test frequency fn, may be understood as the value of the power of signal S at this test frequency fn, divided by the sum of the power of signal S at all test frequencies fn, fn+1, fn+2, fn+3, fn+4, fn+5, fn+6, fn+7, fn+8.
Step 40 is followed by query 41, which checks if there is a danger of feedback. Details pertaining to this query are explained with reference to
If signal S generated by microphone 30 is not already reduced by the bandpass filter, by signal components around the test frequency, then query 42 is followed by a step 43, in which the filter parameters, i.e., mid-frequency fc and quality Q of the bandpass filter, are generated. Mid-frequency fc is an example of the stop frequency along the lines of the claims. The stop frequency may also be, in particular, the frequency range around mid-frequency fc, which the bandpass filter actually filters out from signal S produced by microphone 30.
In the process, mid-frequency fc may be equated with the test frequency. In an example embodiment of the present invention, however, mid-frequency fc is the test frequency, to which a correction frequency is added and at which the power of the signal generated by the microphone is at its maximum; i.e., a correction frequency is added to the test frequency at which the power of the signal generated by the microphone is at its maximum. This correction frequency may be formed as a function of the power of the signal generated by the microphone at the test frequency at which the power of the signal generated by the microphone is at its maximum, as well as a function of the power of the signal generated by the microphone at at least one test frequency existing next to this test frequency. Thus, the correction frequency may be generated in accordance with:
fkorr=sign*fdist*Pmaxneigh/(Pmax+Pmaxneigh), in which:
In the present exemplary embodiment, the correction frequency is formed in accordance with:
fkorr=Δf*(Pneighright−Pneighleft)/(Pmax+|Pneighright−Pneighleft|), in which:
Quality Q is adjusted to a predefined value of, for example, 1/40 Hz.
Step 43 is followed by query 45, as to whether the program is to be terminated. If the program is not to be terminated, then query 45 is followed by step 40. Otherwise, the program is ended.
If signal S generated by microphone 30 is already reduced by the bandpass filter, by signal components around the test frequency, then query 43 is followed by a step 44 in which quality Q is reduced. In this manner, the bandpass filter is adjusted so that it blocks the component of the signal generated by the microphone at an expanded frequency range around mid-frequency fc. Step 44 is followed by step 40.
Provided that there is no danger of feedback, query 41 is followed by query 45 or optionally by a step 46 in which the filtering of signal S generated by microphone 30, around the test frequency, is ended.
An example embodiment of the present invention provides for query 41 to be repeated, at the earliest following a minimum response or dead time, in the present exemplary embodiment, the minimum response time being, e.g., 200 ms to 300 ms.
P(fH0)−P(fH1)≧Δ1
Provision may optionally be made, to supplement query 50 by one or more of the queries:
P(fH0)−P(fH1/2)≧Δ1
P(fH0)−P(fH2)≧Δ1
P(fH0)−P(fH3)≧Δ1
P(fH0)−P(fH4)≧Δ1
it being possible, as the case may be, for other limiting values to be selected, as well.
Test frequencies fn, fn+1, fn+2, fn+3, fn+4, fn+5, fn+6, fn+7, fn+8 illustrated in
If the power of signal S generated by microphone 30 at the test frequency is greater, by not less than a lower limiting value Δ1, than the power of signal S generated by microphone 30 at the first harmonic of the test frequency, then query 50 is followed by a query 51. Query 51 checks whether the power of signal S generated by microphone 30 at the test frequency is greater, by not less than an upper limiting value Δ2, than the power of signal S generated by microphone 30, at the first harmonic of the test frequency. Upper limiting value Δ2 is between 20 and 40 dB, for example. Upper limiting value Δ2 may amount to, e.g., approximately 30 dB. This query is illustrated, by example, in
P(fH0)−P(fH1)≧Δ2
Provision may optionally be made, to supplement query 51 by one or more of the queries:
P(fH0)−P(fH1/2)≧Δ2
P(fH0)−P(fH2)≧Δ2
P(fH0)−P(fH3)≧Δ2
P(fH0)−P(fH4)≧Δ2
it being possible, as the case may be, for other limiting values to be selected, as well.
If the power of signal S generated by microphone 30 at the test frequency is greater, by not more than an upper limiting value Δ2, than the power of signal S generated by microphone 30 at the first harmonic of the test frequency, then query 51 is followed by a query 52, which, by comparing the power of signal S generated by microphone 30 at the test frequency, to the power of signal S generated by microphone 30 at the test frequency at at least an earlier point in time, checks whether the power of the signal generated by the microphone is increasing exponentially at the test frequency.
Provided that there is no danger of feedback or that feedback is not ascertained, query 82 is followed by a query 83 corresponding to query 45 as to whether the program is to be terminated. If the program is not to be terminated, then query 93 is followed by step 81. Otherwise, the program is ended.
If there is a danger of feedback, query 82 is followed by a query 83 corresponding to 42, as to whether signal S generated by microphone 30 has already been reduced by the bandpass filter by signal components around the test frequency. If signal S generated by microphone 30 is already reduced by the bandpass filter, by signal components around the test frequency, then query 83 is followed by a query 85, or alternatively by a query 84.
Query 84 queries as to whether a notch filter is available. If a notch filter is available, query 84 is followed by a step 88, which corresponds to step 43 and in which filter parameters, i.e., for the exemplary embodiment, mid-frequency fc and quality Q of the bandpass filter, are produced. If, on the other hand, query 84 reveals that no notch filter is available, then query 84 is followed by a step 86 in which the power of signal S is reduced by a reduction factor, which may be between, e.g., 2 dB and 5 dB, for the most part, e.g., at 3 dB. Step 86 is followed by a step 87 in which the entire cycle is halted for a pause time of, e.g., approximately 3 s. However, this step may be executed only once per cycle.
Query 85 checks whether a further expansion of the frequency range in which the bandpass filter is blocking, thus a further reduction in its quality Q, would provide that a predefined minimal quality may not be attained. If further expanding the frequency range provides that a predefined minimal quality may not be attained, then query 85 is followed by a step 89, or alternatively by a step 91. In step 91 which corresponds to step 44, quality Q is reduced.
Steps 87, 88 and 91 are followed by a step 92 in which the sequence is paused for a minimum response or dead time, the minimum response or dead time in the present exemplary embodiment being, e.g., 100 ms.
In step 89, the power of signal S is reduced by a reduction factor, which may be between, e.g., 2 dB and 5 dB, for the most part, e.g., at 3 dB. Step 89 is followed by a step 90 in which the entire cycle is halted for a pause time of, e.g., approximately 3 s.
Query 96 checks whether the power of signal S generated by microphone 30 at the test frequency is greater, by not less than 12 dB, than the power of signal S generated by microphone 30, at the first harmonic of (thus twice) the test frequency. If the power of signal S generated by microphone 30 at the test frequency is greater, by not less than 12 dB, than the power of signal S generated by microphone 30 at the first harmonic of the test frequency, then query 96 is followed by a query 97. Otherwise, query 96 is followed by query 93.
A query 97 checks whether the power of signal S generated by microphone 30 is greater at the test frequency, for longer than 750 ms, than a response threshold. If the power of signal S generated by microphone 30 is greater at the test frequency, for longer than 750 ms, than a response threshold, then query 97 is followed by query 83. Otherwise, query 95 is followed by query 93.
The feedback detection in accordance with the present invention is not limited to the example embodiments illustrated in
| Number | Name | Date | Kind |
|---|---|---|---|
| 5574824 | Slyh et al. | Nov 1996 | A |
| 5615302 | McEachern | Mar 1997 | A |
| 5638074 | Johnson | Jun 1997 | A |
| 5872531 | Johnson et al. | Feb 1999 | A |
| 5920840 | Satyamurti et al. | Jul 1999 | A |
| 6674865 | Venkatesh et al. | Jan 2004 | B1 |
| Number | Date | Country |
|---|---|---|
| 39 25 589 | Feb 1991 | DE |
| 41 06 405 | Sep 1991 | DE |
| 197 05 471 | Jul 1997 | DE |
| 199 58 836 | May 2001 | DE |
| 0 078 014 | May 1983 | EP |
| 0 903 726 | Mar 1999 | EP |
| 1 077 013 | Nov 1999 | EP |
| WO 9734290 | Sep 1997 | WO |
| WO 9856208 | Dec 1998 | WO |
| WO 0221817 | Mar 2002 | WO |
| WO 0232356 | Apr 2002 | WO |
| WO 02069487 | Sep 2002 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20040158460 A1 | Aug 2004 | US |
