EP3490270A1 - Method for operating a hearing aid - Google Patents

Abstract

The invention relates to a method for operating a hearing aid (2), wherein a first direction-dependent signal (8f) and a second direction-dependent signal (8r) are generated in the hearing aid (2) from a sound signal (22) from the environment, with the aid of the first direction-dependent signal Signal (8f) and the second direction-dependent signal (8r) a parameter (36) is determined, which represents a quantitative measure for a stationarity of the sound signal (22), and wherein from the first direction-dependent signal (8f) and the second direction-dependent signal ( 8r) a noise-optimized signal (32) is generated on the basis of the parameter (36). The invention also relates to a hearing aid (2) with a first microphone (24a) and a second microphone (24b) for generating a first direction-dependent signal (8f) and a second direction-dependent signal (8r), which is set up to carry out a corresponding method ,

Description

The invention relates to a method for operating a hearing aid, wherein a first direction-dependent signal and a second direction-dependent signal are generated in the hearing aid from a sound signal of the environment, and wherein a noise-optimized signal is generated from the first direction-dependent signal and the second direction-dependent signal.

In hearing aids, one of the most common problems is to improve the signal-to-noise ratio (SNR) for certain listening situations. This is often achieved using directional signal processing algorithms. In this case, it is often assumed that a strongly localized wanted signal component is present in the sound signal of the environment which enters the hearing aid, for example in the form of conversation contributions by a conversation partner. This useful signal component is now delimited by means of direction-dependent signals in the hearing device with respect to a background assumed to be a noise signal, although the noise signal may also have a significant directional dependence. In general, the algorithms mentioned often use self-optimization, wherein the directional characteristic of a direction-dependent signal is adapted in such a way that the influence of interference signals from the direction in which their contribution is greatest is minimized. This is usually done by minimizing the signal power of a corresponding directional signal.

In a first order differential directional microphone with only one adaptation coefficient, often a directional output signal is produced by a linear combination of a forward directed cardioid with a backward directional one Cardioid achieved. A change in the directional characteristic can be achieved via the adaptation coefficient, which determines the contribution of the backward cardioid. As a result, the contributions of noise sources, which may be in a wide solid angle range with respect to the forward direction of the hearing aid, can be reduced. The adaptation is often done so that the energy of the output signal is minimized, since it is assumed that the wearer of the hearing aid aligns its line of sight to the useful signal source, which is represented by the forward Kardioid with constant signal component of the output signal, and thus from other directions incident signals should be noise that is suppressed via the corresponding portion of the reverse cardioid.

If, however, a useful signal does not strike from the forward direction, for example, speech contributions from a speaker positioned laterally to the carrier, these are mitigated accordingly.

The invention is therefore based on the object of specifying a method for operating a hearing device, by means of which a noise can be suppressed with the least possible interference by a useful signal, regardless of its direction.

The object is achieved by a method for operating a hearing aid, wherein in the hearing aid from a sound signal the environment, a first directional signal and a second directional signal are generated, wherein a parameter is determined based on the first directional signal and the second directional signal, which represents a quantitative measure of a stationarity of the sound signal, wherein from the first directional signal and the second directional signal based on the parameter, a noise-optimized signal is generated, and wherein the parameter is determined in a signal feedback loop from the noise-optimized signal. Advantageous and partly inventive in themselves embodiments are the subject of the dependent claims and the following description. The first direction-dependent signal and the second direction-dependent signal are preferably each generated on the basis of corresponding signals from at least two input sound transducers, which may be provided by microphones, for example. In this case, a direction-dependent signal in each case means a signal which has a non-trivial directional characteristic, ie, for a test sound with a constant sound pressure and the corresponding test sound source at a constant distance from the hearing aid, the sensitivity to the test sound in the respective direction-dependent signal has a measurable, prefers considerable directional and in particular angle dependence in the transverse plane of the carrier.

In this case, a quantitative measure of a stationarity is in particular a measure which assigns a numerical value to a signal in such a way that the extreme value of the measure is assumed for a pure sine tone of constant frequency, and a correspondingly monotonous change with an increasing variation of frequencies of signal components listed in the figure. Preference can be given to those skilled in common definitions of stationarity for the above assignment. The parameter can represent an absolute quantitative measure, which measures the stationarity of the signals to be checked on the basis of a normalized scale, and in particular have a fixed maximum and a fixed minimum value, or a relative measure, which in particular no fixed extremal value for non-stationary Has signals.

In particular, a noise-optimized signal comprises a signal which, with respect to the useful signal components contained in the sound signal, has an SNR optimized with respect to the first direction-dependent signal and with respect to the second direction-dependent signal, if the useful signal components in the sound signal are superposed by noise components. In particular, the first direction-dependent signal and the second direction-dependent signal can enter the noise-optimized signal linearly, ie, the noise-optimized signal with respect to one at a particular time in one of the two direction-dependent Signaling change in the time-frequency domain has a linear response.

An approach to noise reduction common in hearing aids is first to design the first directional signal so that its direction of maximum sensitivity coincides with the wearer's frontal direction. The second directional signal is now designed to show, on one hand, its direction of maximum sensitivity in a direction other than the frontal direction of the carrier, and instead the direction of minimum sensitivity coincides with the frontal direction of the carrier. For the directional characteristic of the second direction-dependent signal, the directional characteristic of the first direction-dependent signal with respect to the frontal plane of the carrier is preferably mirrored during proper wear during operation.

To suppress a noise, the first direction-dependent signal, which primarily receives the speech signal components of a conversation partner in the frontal direction, is superimposed on the second direction-dependent signal as a function of the total energy of a resulting signal. In this case, by the second direction-dependent signal signal components, which do not impinge on the carrier from the frontal direction, and are thus assumed to be noise, can be suppressed. Due to the constant contribution of the first directional signal in the frontal direction is required for effective suppression of noise only the said condition of the minimum total energy of the resulting signal from the superposition.

In contrast, it is now proposed to investigate the stationarity of the sound signal via a corresponding parameter on the basis of the first direction-dependent signal and the second direction-dependent signal. The procedure proposed in the invention is based on the consideration that strongly directed interference noise, as may be the case, for example, due to the hum of an engine or a domestic appliance, is satisfactorily suppressed by the previous procedure in the event of a lateral impact on the carrier can be, however, in the event that the signal striking the side is a useful signal, so for example, a voice signal of another speaker to be added, also a suppression takes place, which would be undesirable in this case. For this purpose, a distinction between a possible useful signal and a possible noise is performed, taking into account that usual useful signals such as speech or music usually have a much lower stationarity than most directional noise and as a diffuse background noise, such as in a conversation can occur with several people in a room, in which further conversations take place (so-called "cocktail party" listening situation).

This now makes it possible, for example, to generate the noise-optimized signal from the two direction-dependent signals at a low stationarity in such a way that the lowest possible directional suppression of signal components takes place and thus any speech signals impinging laterally on the carrier are not suppressed, but amplified. In turn, at a detected increased stationarity assuming that now significant noise should be present, a directional suppression carried out such that the noise-optimized signal vorranging includes only the voice signal of a conversation partner, on which out preferred to align the direction of maximum sensitivity of the first directional signal is.

According to the invention, the parameter is determined in a signal feedback loop from the noise-optimized signal. While the parameter could also be determined technically from the first direction-dependent signal and the second direction-dependent signal-that is to say without further processing for the noise-optimized signal itself-a determination of the parameter from the noise-optimized signal has the advantage that this signal is for further processing in the hearing device provided signal is, can be used as a target size. Elaborate conversions can thus be omitted.

It proves to be advantageous if an autocorrelation function is determined as a parameter. In this case, the autocorrelation function is preferred over one to determine the expected useful signals and the expected noise appropriate time window to be determined. The advantage of using the autocorrelation function as a parameter is that it often provides additional valuable information that may be of concern in subsequent signal processing.

Conveniently, the noise-optimized signal is generated by a superimposition of the first direction-dependent signal and the second direction-dependent signal, wherein a weighting factor for the superimposition is determined on the basis of the parameter. This means, in particular, that the noise-optimized signal is of the form F + α * B, where F denotes the first directional signal and B the second directional signal, and α is the weighting factor determined by the parameter. On the one hand, this superimposition is technically particularly easy to implement; on the other hand, the first direction-dependent signal can be aligned in such a way that the direction of maximum sensitivity is directed toward a support partner of the wearer, in particular in the frontal direction, which further facilitates the determination of the weighting factor α.

In this case, the noise-optimized signal for a non-stationary sound signal preferably has an essentially omnidirectional directional characteristic due to the weighting factor, and a maximally directional directional characteristic for a maximally stationary sound signal as a consequence of the weighting factor. In this case, a maximum directional directivity is to be understood as meaning, in particular, a global maximum of the directivity within the scope of the available directional signals. This takes into account the fact that it is assumed for non-stationary sound signals that there is no noise to be suppressed, but that, on the other hand, speech signals impinging laterally on the carrier may be present. In this case, a substantially omnidirectional directional characteristic of the noise-optimized signal is advantageous, since in this way speech signals from all spatial directions can be taken into account. In return, it is assumed for a maximum stationary sound signal that a significant amount of noise is present, which is correspondingly by a Directional characteristic of the noise-optimized signal is to suppress such that only the spatial direction in which a conversation partner is assumed, so usually the frontal direction, provides significant signal contributions to the noise-optimized signal. In this case, a substantially omnidirectional directional characteristic is to be understood as meaning, in particular, such a directional characteristic in which a deviation from a perfect omnidirectionality with respect to the directivities occurring, in particular in the case of the directional directional characteristics, can be neglected.

In a further advantageous embodiment of the invention, the parameter is determined in such a way that the noise-optimized signal is minimal in terms of the parameter. This can be done in particular by minimizing the noise-optimized signal with respect to the parameter. This procedure has the advantage that the noise-optimized signal always has the lowest possible stationarity and thus always the lowest possible noise component.

The noise-optimized signal is expediently minimized with regard to its signal energy and with regard to the parameter. This means, in particular, that the noise-optimized signal, which is formed from the first direction-dependent signal and the second direction-dependent signal, has a local minimum as a function of the variables signal energy and of the parameter. In this way, it is possible in particular to suppress such interference noises which impinge on the carrier from different directions in a complex auditory situation, wherein, moreover, a diffuse noise background may be present, while the interfering noises themselves may only partially be assumed to be stationary.

Conveniently, in this case the first direction-dependent signal and / or the second direction-dependent signal are generated on the basis of a time-delayed superposition of the first microphone signal with the second microphone signal. Preferably, the acoustic transit time difference between the first microphone and the second microphone is used for the time delay in the overlay. This is a very easy to implement yet efficient process for generating a directional signal when the underlying microphone signals are from directional microphones.

In this case, the first direction-dependent signal particularly preferably has a directional dependence in the form of a first cardioid, which is aligned in a first direction, and / or the second direction-dependent signal has a directional dependence in the form of a second cardioid, which is aligned in a second direction. A cardioid-shaped signal is characterized by the fact that the direction of minimum sensitivity is opposite to the direction of maximum sensitivity. This is not the case, for example, for signals whose directional characteristic forms a supercardioid or a hypercardioid. In addition, a sound signal from the direction of minimum sensitivity is ideally completely suppressed in a cardioid-shaped directional characteristic. The symmetry between the direction of maximum and minimum sensitivity thus makes it possible to keep calculations for the first and the second superimposition for noise suppression particularly simple, since in addition a strictly monotonous increase in sensitivity takes place from the direction of minimum sensitivity to the direction of maximum sensitivity , Particularly preferably, the first direction of the second direction is opposite in this case.

Against the background that in a directional signal with cardioid-shaped directivity sound signals from the direction of the minimum sensitivity are ideally completely suppressed, thereby the calculation of the concrete weighting of the two directional signals in the overlay can be further simplified because the first directional signal can be taken as a reference directed to a first useful signal source, and in this case, when the second cardioid-shaped directional signal is aligned against the first directional signal, noise suppression by the second directional signal has no effect on the signal Contribution of the first useful signal has.

Thus, to determine the weights for the most efficient noise suppression possible in the case of stationary signals, simply a minimum signal power be required in the resulting signal from the superposition, without this having an impact on the contribution of the first useful signal. For this purpose, the superimposition is preferably initially formed on the basis of the minimum signal power, then the quantitative parameter for the stationarity is determined for the resulting signal, and the weighting in the superimposition is then adjusted on the basis of the parameter, in particular iteratively, until the parameter is minimal the resulting signal has a minimum stationarity relative to the parameter.

The invention further mentions a hearing aid having a first microphone and a second microphone for generating a first direction-dependent signal and a second direction-dependent signal, which is set up to carry out the method described above. In particular, in this case the first direction-dependent signal and the second direction-dependent signal are respectively generated by means of both the first microphone and the second microphone. Preferably, the method is carried out during operation of the hearing aid by means of a control unit, which is particularly preferably designed as part of the signal processing unit in which all other functions of the signal processing are implemented. The advantages stated for the method and for its further developments can be transferred analogously to the hearing aid.

Fig. 1

in einer Draufsicht die Abschwächung eines gerichteten Störsignals mittels einer Überlagerung zweier Richtsignale in einem Hörgerät, und

Fig. 2

in einem Blockdiagramm den Ablauf eines Verfahrens zur Abschwächung von Störsignalen in einem Hörgerät bei gleichzeitiger Anwesenheit von Nutzsignalen aus unterschiedlichen Richtungen.

An embodiment of the invention will be explained in more detail with reference to a drawing. Here are shown schematically in each case:

Fig. 1

in a plan view, the attenuation of a directional interference signal by means of a superposition of two directional signals in a hearing aid, and

Fig. 2

in a block diagram, the sequence of a method for attenuation of interference in a hearing aid with simultaneous presence of useful signals from different directions.

Corresponding parts and sizes are provided in all figures with the same reference numerals.

In Fig. 1 schematically a carrier 1 of a hearing aid 2 is shown in a plan view. The carrier 1 is here in a conversation situation with a conversation partner 4, which is positioned with respect to the carrier 1 in the frontal direction 6. In a manner not shown, a first direction-dependent signal 8f (dashed line) and a second direction-dependent signal 8r (dotted line) are now formed in the hearing aid 2, the directional characteristic is given in each case by a cardioid. The Kardioid-shaped directional characteristic of the first directional signal 8f has the consequence that for sound signals from the frontal direction 6 is a maximum sensitivity and thus received sound signals from this direction maximum in the first directional signal 8f, while sound signals from the frontal direction 6 opposite reverse direction 10th ideally be completely suppressed in the first directional signal 8f. The second direction-dependent signal 8r has a direction dependence opposite to the first direction-dependent signal 8f, so that sound signals from the reverse direction 10 are received in the second direction-dependent signal 8r maximally, while sound signals from the frontal direction 6 are ideally completely suppressed.

The conversation situation of the wearer 1 with the conversation partner 4 is now superimposed here by various noise 12a, 12b, wherein 12a and 12b are highly directional noise, that is, each of a localizable source such. a motor or a household electrical appliance are emitted.

To correct the speech signal 13 of the conversation partner 4 from the noise 12a, 12b are now attenuated in the hearing aid 2 by a superposition of the first directional signal 8f with the second directional signal 8r the form F + α · B, where F and B, the first or second direction-dependent signals 8f, 8r and α is a weighting factor to be selected accordingly. This exploits that the Nutzsignalquelle, so here the caller 4, as is assumed in the frontal direction 6, and thus their contributions in the second directional signal 8r are completely suppressed, and therefore only by the first directional signal 8f in the superimposing resulting signal F + α B input. The contribution of the second direction-dependent signal 8r is therefore to be adjusted in the resulting signal via the weighting factor α such that the resulting signal has a minimal signal level, not least as a result of the constant input of the useful signal from the frontal direction 6 (see above) given a variation of .alpha is made that the attenuation of the signal components, which do not come from the frontal direction 6, is maximum.

If a further speaker 14 now appears whose voice signal 16 does not impinge on the carrier 1 from the frontal direction 6, but rather from a lateral direction, the procedure just described would initially ensure that the voice signal 16 is treated like the noise 12a, 12b and accordingly is suppressed. In order to avoid this, a detection is made as to whether the laterally impinging noises 12a, 12b, 16 are noises or potential useful signals, and only suppress the noises 12a, 12b. This is based on Fig. 2 described.

In Fig. 2 a method 20 for directional noise suppression in the hearing aid 2 is shown by means of a block diagram. In the hearing device 2, a first microphone signal 26a is generated from the sound signal 22 of the environment by means of a first microphone 24a, and a second microphone signal 26b is generated by means of a second microphone 24b. The second microphone signal 26b is delayed on the one hand by the time interval T, so that thereby a time-delayed second microphone signal 28b is formed, which is subtracted from the first microphone signal 26a, so that thereby the first directional signal 8f is formed. In the same way, the first microphone signal 26a is additionally delayed by the time interval T, thereby forming the first time-delayed microphone signal 28a, which is subtracted from the second microphone signal 26b, thereby forming the second direction-dependent signal 8r. The first directional Signal 8f and the second direction-dependent signal 8r here respectively exhibit the cardioid-shaped directional characteristics Fig. 1 on.

In a superimposition 30 of the form F + α * B, a weighting factor α is now determined in such a way that the signal 32 resulting from the superimposition 30 has a minimal stationarity. For this purpose, the resulting signal 32 is supplied to a signal feedback loop 34, where it determines a parameter 36 for the stationarity of the signal components. The parameter 36 may for example be given by an autocorrelation function, which is to be calculated over a time window to be suitably selected.

If it is then found that the signal 32 resulting in a weighting factor α has a minimum stationarity with respect to the parameter 36, ie the parameter 36 for the present overlay 30 assumes a local minimum, then the overlay 30 is formed in the signal feedback loop 34 not changed. However, if it is determined in the signal feedback loop 34 that the resulting signal 32 has a stationarity parameter 36 which is not minimal, e.g. by considering the monotony of the parameter 36 with small variations of α around the present value, the weighting factor α in the overlay 30 is adjusted to minimize the parameter 36. This can be done interactively in particular. Also conceivable is a parameter 36 which provides an absolute measure of stationarity and is in particular suitably normalized, so that on the basis of the value parameter 36 to a present superimposition with a weighting factor α and based on the corresponding distance of the value of the parameter 36 from the minimum value also a quantitative statement about the necessary adaptation of the weighting factor α is possible.

Lying for example in the conversation situation Fig. 1 only the speech signal 13 of the interlocutor 4 and the two noise noises 12a, 12b, the speech signal is non-stationary, while the two noises 12a, 12b are highly stationary. For superposition 30, the weighting factor α is to be determined in such a way that in F + α * B the signal components of the interference noises 12a, 12b are eliminated via the second directional signal B as possible. This is done by a negative weighting factor α of the amount <1. In this case, the resulting signal 32 substantially corresponds to that signal which would also be achieved by minimizing the signal energy, since the speech signal 13, which enters the resulting signal 32 through F, is non-stationary, and its signal components through the corrections the stationary signal components are not affected by the signal B.

On the other hand lie in the conversation situation Fig. 1 only the speech signal 13 of the conversation partner 4 and the speech signal 16 of the conversation partner 14 before, would an overlay 30, which is based on minimizing the energy of the resulting signal 32, suppress the speech signal 16 of the conversation partner 14 significantly, which is undesirable. However, since the determination of the weighting factor α is not based on minimizing the energy of the resulting signal 32, but on minimizing its stationarity - as measured by the parameter 36 - the signals are largely added in F + α * B, resulting in a largely omnidirectional directional characteristic for the resulting signal 32.

Due to the additional contributions of the speech signal 14 in the signal B, the already low stationarity of the speech signal 13 as a result of the different call partners 4, 14 and thus the different spectral contributions is further reduced in the resulting signal 32. The weighting factor α is now positive, and designed so that it compensates as possible for the attenuation of the speech signal 16 by the lateral attenuation of the directivity of the first directional signal 8f.

If in the conversation situation after Fig. 1 If both speech signals 13, 16 and both noise 12a, 12b are present, the minimization of the stationarity of the resulting signal 32 will result in the one hand, the stationary noise 12a, 12b provide the lowest possible contributions to the resulting signal 32, on the other hand, the non-stationary speech signal 16 as possible little is suppressed. Since only one degree of freedom - the weighting factor α - is available, is this is only possible with restrictions, the resulting signal 32 is no longer minimal in terms of the signal energy in a, but this is accepted in view of the complex conversation situation in order not to suppress the speech signal 16 undesirably.

By the procedure described on the one hand noise of the form 12a, 12b suppressed, however, not the signal components of the speech signal 16 so that the resulting from the superposition signal 32 is a noise-optimized signal.

Although the invention has been illustrated and described in detail by the preferred embodiment, the invention is not limited by this embodiment. Other variations can be deduced therefrom by those skilled in the art without departing from the scope of the invention.

LIST OF REFERENCE NUMBERS

1

carrier

2

hearing Aid

4

interlocutor

6

frontal direction

8f

first directional signal

8r

second directional signal

10

reverse direction

12a, b

interference

13

speech signal

14

interlocutor

16

speech signal

20

method

22

sound signal

24a / b

first / second microphone

26a / b

first / second microphone signal

28a / b

first / second time delayed microphone signal

30

overlay

32

resulting / noise optimized signal

34

Signal feedback loop

36

parameter

T

time interval

Claims (11)

Method for operating a hearing aid (2), wherein a first direction-dependent signal (8f) and a second direction-dependent signal (8r) are generated in the hearing device (2) from a sound signal (22) of the surroundings, wherein, based on the first direction-dependent signal (8f) and the second direction-dependent signal (8r), a parameter (36) is determined which represents a quantitative measure of stationarity of the sound signal (22), - Wherein a noise-optimized signal (32) is generated from the first direction-dependent signal (8f) and the second direction-dependent signal (8r) on the basis of the parameter (36), and - Wherein the parameter (36) in a signal feedback loop (34) from the noise-optimized signal (32) is determined. Method according to claim 1,
wherein an autocorrelation function is determined as parameter (36). A method according to claim 1 or claim 2,
wherein the noise-optimized signal (32) is generated by a superposition of the first direction-dependent signal (8f) and the second direction-dependent signal (8r), and
wherein a weighting factor for the overlay is determined using the parameter (36). Method according to claim 3,
wherein for a non-stationary sound signal (13, 16) by the weighting factor, the noise-optimized signal (32) has a substantially omnidirectional directivity, and
wherein for a maximum stationary sound signal (12a, 12b) by the weighting factor, the noise-optimized signal (32) has a maximum directional directivity. Method according to one of the preceding claims,
wherein the parameter (36) is determined such that the noise optimized signal (32) is minimal with respect to the parameter (36). Method according to one of the preceding claims,
wherein noise optimized signal (32) is minimized with respect to its signal energy and parameter (36). Method according to one of the preceding claims,
wherein in the hearing aid (2) from the sound signal (22) by a first microphone (24a), a first microphone signal (26a) is generated, and by a second microphone (24b), a second microphone signal (26b) is generated, and
wherein the first directional signal (8f) and / or the second directional signal (8r) are generated from the first microphone signal (26a) and the second microphone signal (26b). Method according to claim 7,
wherein the first directional signal (8f) and / or the second directional signal (8r) are generated by means of a time-delayed superposition of the first microphone signal (26a) with the second microphone signal (26b). Method according to claim 8,
wherein the first directional signal (8f) has directionality in the form of a first cardioid aligned in a first direction (6), and / or
wherein the second directional signal (8r) has a directional dependence in the form of a second cardioid aligned in a second direction (10). Method according to claim 9,
wherein the first direction (6) is opposite to the second direction (10). Hearing aid (2) with a first microphone (24a) and a second microphone (24b) for generating a first direction-dependent signal (8f) and a second direction-dependent signal (8r), which is adapted to carry out the method according to one of the preceding claims.

Images