EP3598778A1 - Hearing device and method for operating a hearing device for detecting the own voice on the basis of an individual threshold value - Google Patents

Abstract

The invention relates to a method for operating a hearing device (2), whereby a noise is recorded by means of a microphone (4), the noise being analyzed with regard to its correspondence with the hearing aid wearer's own voice and a characteristic value (M) which indicates how closely the noise corresponds to the hearing aid wearer's own voice, the own voice being a noise type (G1), the feature value (M) being compared with a threshold value (S), the noise depending on whether the feature value (M) is above or below the threshold value (S), is recognized as a separate voice, and the hearing device (2) is switched between several operating modes depending on whether the sound was recognized as a separate voice. The method is characterized in that the threshold value (S) is set depending on the user. As a result, an improved self-voice recognition (10) is formed, which distinguishes the hearing aid wearer's own voice from another type of noise (G2) particularly reliably. The invention further relates to a hearing aid (2) with a corresponding natural voice recognition (10).

Description

The invention relates to methods of operating a hearing aid, wherein a noise is recorded by means of a microphone, the noise being analyzed with regard to its correspondence with the hearing aid wearer's own voice and a feature value being generated which indicates how strongly the noise with the hearing aid wearer's own voice matches, the own voice being a type of noise, the feature value being compared with a threshold value, the noise being recognized as a separate voice depending on whether the feature value is above or below the threshold value, and the hearing device depending on whether the sound was recognized as a separate voice, is switched between several operating modes. The invention further relates to a hearing aid.

A corresponding procedure is, for example, in the unpublished, international application attributed to the applicant with the file number PCT / EP 2015/068796 described.

In the context of an analysis of the noises, which are recorded by means of one or more microphones, it is possible to recognize the hearing aid wearer's own voice and, depending on this, to switch the hearing aid between different operating modes. Such an analysis is also referred to as self-voice detection or "own voice detection", or OVD for short, and is carried out by means of self-voice recognition, which is usually a component of the hearing aid. The noise is converted into electrical signals by the microphone, which are then examined in order to assign the noise to a specific type of noise, more precisely to decide whether whether the original sound is your own voice or not, ie whether the hearing aid wearer speaks or not.

From the US 2011/0261983 A1 A method for recognizing one's own voice can be found, in which a predetermined threshold value is selected for recognizing one's own voice as a function of ambient noise. For this purpose, different threshold values are initially defined for different noise classes in ambient noise. During normal operation, ie when the hearing aid is used by the hearing aid wearer, the threshold value is selected as a function of the current noise class.

In the application cited at the beginning PCT / EP 2015/068796 the analysis is carried out using special filters, each of which has its own filter profile that is adapted to a particular noise, ie to a specific noise type or a specific noise class. A given signal is then filtered by means of the filter. From the resulting filtered signal, it is then determined for each of the filters how much the original noise corresponds to the type of noise to which a respective filter is adapted. For this purpose, the filter profiles are designed, for example, in such a way that the noise to be detected is maximally attenuated due to the filter profile. In the aforementioned application, a distinction is made according to the location of the noise, ie noises that occur at different points in space relative to the hearing aid are influenced differently by a respective filter. A spatial differentiation is possible and a differentiation of the noise type based on its position relative to the hearing aid. So nearby sounds are recognized as spatially close and then accepted as their own voice, while sounds further away are recognized as such and then accepted as a foreign voice. A greater agreement of the actual noise with the noise to which the filter is adapted leads to a greater attenuation and a higher agreement, ie to a higher probability that the examined noise corresponds to the noise type which is assigned to the filter. In this way, noise can be correct with a certain probability classify and assign one of in particular several different sound types.

Applying different filters to a recorded signal results in different values for the attenuation, i.e. generally the agreement, so that based on these values it can be decided which type of noise it is. If the hearing device wearer now speaks himself, then the signal is attenuated more strongly by this filter and there is a higher value for the match than with another filter, which is adapted, for example, to a foreign speaker in the front area of the hearing device wearer. By evaluating the two values, it can then be reliably determined that the hearing aid wearer is speaking, i.e. there is an own voice situation. The evaluation is carried out by forming a feature value, for example by forming a difference or quotient of the two values for the attenuation, and then comparing the feature value with a predetermined, stored threshold value or limit value.

Against this background, it is an object of the invention to provide a method for operating a hearing device, in which a distinction between the hearing aid wearer's own voice and other noises is made more reliably. Furthermore, a corresponding hearing aid with improved self-voice recognition is to be specified.

The object is achieved according to the invention by a method with the features according to claim 1 and by a hearing device with the features according to claim 11. Advantageous refinements, developments and variants are the subject of the dependent claims. The explanations in connection with the method apply mutatis mutandis to the hearing aid and vice versa.

The method is used to operate a hearing aid. Hearing aid is generally understood to mean a device for outputting sound, ie noise using a loudspeaker, the sound being obtained from noise which was recorded from the surroundings by means of at least one microphone. The noise marriage are converted into electrical signals by the microphone and processed in the hearing aid by means of a control unit. The signals are then converted into noise again via the loudspeaker and output. In particular, a hearing aid is understood to mean a device for the care of a hearing-impaired or hearing-impaired person, which in particular wears the hearing aid continuously or for most of the time in order to compensate for a hearing deficit. The hearing aid thus has a total of at least one microphone, a loudspeaker, also referred to as a receiver, and a control unit, the latter controlling the recording of sounds and their output. The control unit is usually designed at least to amplify noise.

In the method, a noise is picked up by means of the microphone. The noise, more precisely the electrical signal generated therefrom, is analyzed with regard to its correspondence with the hearing aid wearer's own voice and a characteristic value is generated which indicates how closely the noise corresponds to the hearing aid wearer's own voice. The own voice is a type of noise, in particular of several different types of noise.

The feature value is preferably generated using a classifier. A classifier analyzes a recorded noise with regard to a number of characteristic features of a certain noise type and supplies the feature value as a measure of the correspondence with the noise type. The characteristic value is then compared with a threshold value. Depending on whether the feature value is above or below the threshold value, the noise is recognized as a separate voice, i.e. clearly assigned to the sound type "own voice". In this respect, the comparison with the threshold value is a decision-making process for determining which values of the characteristic value are based on the presence of one's own voice and when in the end your own voice is considered to be recognized.

The analysis of the noise, the generation of the characteristic value, the comparison with the threshold value and the decision whether the own voice is present or not, are carried out by means of natural voice recognition, which is a component of the hearing aid and which is implemented, for example, as an integrated circuit. The natural voice recognition can be part of the control unit of the hearing device or can be designed as a separate unit. Depending on whether the noise has been recognized as a separate voice, the hearing device is switched between several operating modes, for example an individual voice mode and a non-individual voice mode. The switchover takes place automatically, that is to say by the hearing aid itself, in particular by the control unit or directly by the natural voice recognition.

According to the invention, the threshold value is set as a function of the user and as an individual threshold value.

User-dependent determination of an individual threshold value means that the threshold value is set depending on the person of the hearing device wearer. In particular, no feature values from other hearing device wearers / users are used to determine the threshold value.

The setting is made either as part of a fitting session with the acoustician, by the hearing aid wearer himself or in normal operation, i.e. online, and automatically through the hearing aid. By adapting the threshold value used for comparison to the user, a possibly very different characteristic value is optimally taken into account when determining, in particular classifying, one's own voice. The generation of the feature value itself, as already described above, is also expediently adapted specifically to the hearing device wearer in order to realize a particularly optimal recognition of one's own voice.

For the user-dependent, individual setting, the threshold value is determined by means of a calibration method, in which the hearing aid wearer's own voice is recorded several times and several individual, ie user-specific, characteristic values are generated. Finally, in the calibration process, the individual threshold value becomes dependent on the individual ones generated Characteristic values set. In this way, a particularly suitable and user-optimal threshold value is set. A large number of individual characteristic values are therefore generated, so that a distribution of the individual characteristic values is obtained, from which the threshold value is then determined.

The threshold value is therefore set as a function of the individual feature values generated in the calibration process by setting the threshold value with respect to a characteristic value of the distribution, for example as a 2σ deviation from the mean value or generally in such a way that the generated feature values are predominantly above or below the threshold value.

This configuration is based on the knowledge that the threshold value can be strongly user-dependent. Especially in the case of the one described at the beginning and from the PCT / EP 2015/068796 In the method to be extracted, the attenuation values generated by the filter used can vary greatly depending on the user. A fixed threshold value would therefore result in one user being recognized in their own voice and in the other user being recognized in a foreign voice, although in both cases their own voice is present.

Furthermore, this embodiment is based on the consideration that both the own voice and external voices / ambient noises are recorded over time during the calibration process. Characteristic values are therefore obtained in the presence of one's own voice as well as in the presence of a foreign voice / ambient noise. The distribution of the characteristic values obtained overall therefore shows a range of possible characteristic values. The individual threshold value is determined from this distribution, for example by means of statistical methods, in particular averaging.

This is based in particular on the knowledge that a characteristic value, which is determined and used to identify a noise and to assign it to a noise type, can be very different due to the environment. In other words: in different environments for the hearing aid, a feature value that is sometimes greatly changed may be present of the detection of a certain noise, since this is, for example, changed, distorted or superimposed by other noises. The term “environment” is to be understood in relation to the hearing device and not in relation to the hearing device wearer. In particular, the hearing aid wearer's own voice is logically different from user to user, so that different hearing aid wearers also represent different environments for the hearing aid. However, other noises, ie noises external to the hearing aid wearer, for example external voices, can also lead to different characteristic values in different environments.

"Noise" generally means any kind of sound signals in the audible frequency range. Different types of noise include your own voice, a foreign voice, tones, sounds, music, background noise and noise.

The method according to the invention is further based on the consideration that a decision of the self-voice recognition is potentially highly error-prone due to a predetermined threshold value. In order to reduce the error in determining the noise type of a noise, it is fundamentally possible to intentionally set the threshold value particularly high or particularly low. As a result, the error rate in the erroneous recognition of noises other than one's own voice, or conversely the error rate in the non-recognition of one's own voice, although this is present, can be reduced; overall, however, this approach is inadequate, since the correct recognition or non-recognition of one's own voice is restricted to particularly clear cases, and the particularly environment-dependent range of values of the feature values is thereby largely excluded.

A user-dependent setting of the threshold value is understood in particular to mean that no generally predetermined threshold value is used by the self-voice recognition for decision making.

The suitable threshold value is rather selected in particular by a previous environmental analysis. In this case, for example, the current environment is first appropriately determined by the self-voice recognition itself or by the control unit and then the associated threshold value that is optimal for the environment is selected and set from a group of threshold values.

A distinction must be made in advance from the above-described, environment-dependent setting of the threshold value in operation for the specific threshold value to be used for this particular situation. This determination is made either when adjusting the hearing aid, e.g. as part of a fitting session with the acoustician, or alternatively or additionally by the hearing aid wearer himself. An automatic determination in a special calibration operation or in normal operation of the hearing aid is also fundamentally conceivable. In general, the definition creates an assignment of threshold values to environments, so that a group of threshold values is available for selection, from which the most suitable is then set. This assignment is expediently stored in a memory of the hearing device, in particular the control unit, for example as a table, as a functional assignment or as a user profile. In this respect, not only is a predetermined threshold value stored, but a plurality of predetermined threshold values are stored for different environments. A suitable one is then selected and set depending on the environment from these several predetermined threshold values, as a result of which the selection of the operating mode of the hearing device is significantly less prone to errors during operation.

The user-dependent setting of the individual threshold value is further to be distinguished from setting the determination of a feature value, for example setting the filter mentioned at the beginning or a classifier, which is used for analyzing noises and for generating a feature value. The threshold value is therefore not used to determine the characteristic value, but rather to evaluate the characteristic value already determined. Such a configuration of those components that have the characteristic values generate, rather takes place in particular independently of the user-dependent or environment-dependent selection and setting of the threshold value for the evaluation of the feature value. However, these components are expediently also set as a function of the user. This is useful, for example, with regard to the self-voice recognition, that is to say the recognition of the voice of the hearing device wearer, that is to say the generation of the feature value, for example by means of a filter, is expediently adapted to the voice of the hearing device wearer in order to ensure optimum generation of feature values and thus optimal differentiability from other types of noise.

In a suitable development, the threshold value is calibrated by determining a maximum and a minimum feature value over a limited period of time and by setting the threshold value between the minimum and the maximum feature value. This is based in particular on the assumption that the noise of the noise type is "own voice" at the maximum feature value and that of the noise type is "foreign voice" at the minimum feature value. Depending on the calculation of the feature value, however, this can also be the other way round, ie it is then assumed that one's own voice generates a minimum feature value and the foreign voice generates a maximum feature value. The limited time period is usually from a few seconds to a few tens of seconds, for example about 20 s. The maximum and the minimum characteristic value are therefore short-term extremes within the period. Through the continuous determination of short-term extrema, characteristic values are determined over a clearly longer period than the limited period, on the one hand for one's own voice and on the other hand for another type of noise, in particular a foreign voice. In this way, at least similar statistical distributions are advantageously obtained as in the calibration method mentioned above, in which at least the presence of one's own voice must be known in particular. On the other hand, based on the minimum and maximum feature values within the limited time period, it is advised, in particular, when the recorded sound is your own voice and when it is a different type of sound.

In an advantageous embodiment, the threshold value is calibrated in normal operation by repeatedly determining the individual feature values and setting the threshold value depending on it. As a result, the threshold value is continuously adapted so that the threshold values stored as part of the assignment approach the optimal threshold values over time.

The calibration does not correspond to the environment-dependent setting of the threshold value, which is set in a specific situation. Rather, during calibration, the threshold value stored for a respective value range is adjusted, which is then set. In this sense, the recurring recalibration of the threshold value of a value range is a continuous online optimization of the self-voice recognition. This optimization is carried out either continuously or only at certain times or only over a single certain period.

In an advantageous embodiment, the noise is analyzed in addition to the agreement with one's own voice with regard to agreement with at least one other type of noise. In this case, for example, a match value is generated in each case, which indicates how strongly the noise matches a specific noise type, the match values then being combined to form the feature value. One of the at least two types of sounds is your own voice. A distinction is then made between the own voice and the other type of noise on the basis of the characteristic value. This distinction is significantly improved by the threshold value set according to the environment. The characteristic value is, for example, the difference or the quotient from the two match values.

In a preferred variant, the distinction between one's own voice and another type of noise corresponds to the distinction between locally, ie spatially separated, noises. Your own voice is usually the type of noise that is closest to the hearing aid, so that spatial differentiation, ie differentiation according to the location of the noise, a simple distinction is made between your own voice and another type of noise.

In a preferred development, the other type of noise is a foreign voice, which is arranged in particular frontally with respect to the hearing aid wearer. In this context, a foreign voice is not understood to mean, in particular, the voice of a specific other person, but quite generally a voice which is not the hearing aid wearer's own voice. A distinction is then made between the own voice and a foreign voice by means of the self-voice recognition.

In a particularly preferred embodiment, the feature value is generated as in the international application mentioned at the beginning PCT / EP 2015/068796 by means of a pair of filters, one of the filters being configured for maximum attenuation of one's own voice and the other filter for maximum attenuation of an alien voice, in particular an alien voice, which comes from a person in front of the hearing aid wearer. When analyzing a noise, the two filters each provide a match value and the feature value is then formed from the two match values, for example by subtracting the match value with respect to the foreign voice from that of one's own voice. The characteristic value is then lower with a foreign voice than with one's own voice. If the threshold is undershot, the noise is recognized as a foreign voice; however, if the threshold is exceeded, the noise is recognized as a separate voice.

The generation of the characteristic values is also user-dependent for other types of noise. In an advantageous further development, the calibration method therefore records another type of noise, in particular a foreign voice, before or after recording one's own voice. Here too, in particular analogously to what has been said, several feature values are generated, depending on which of the threshold values is set. The calibration is thus significantly improved, especially with regard to the accuracy in the differentiation between your own voice and the other type of noise. The mean value of the two mean values of the two statistical distributions generated for the two types of noise is then set as the threshold value.

The person of the hearing aid wearer is not the only environmental condition with respect to which the threshold value is usefully adjusted. Of particular importance in the analysis of most types of noise is their superposition with a noise, often a background noise or a background noise. In particular, it was recognized that the generation of a feature value, i.e. in particular the classification of the noise, with increasing volume of the noise becomes more difficult and more error-prone. The same applies analogously to the distinction between two types of noise. Therefore, in a particularly preferred embodiment and as an alternative or in addition to the user-dependent setting of the threshold value, the threshold value is set depending on the environment in that a noise value is determined and the threshold value is set as a function of the noise value. This further optimizes the natural voice recognition.

The noise value characterizes the noise and quantifies it in particular. The noise value is preferably a level, a volume, an intensity or an amplitude of the noise. Alternatively, the signal-to-noise ratio is also suitable as a noise value. A typing of the noise is also suitable, i.e. the assignment of the current noise to a specific noise type and a setting of the threshold value depending on the detected noise type, the noise type then being the noise value.

In addition or as an alternative to the noise-dependent setting, any other environmental dependency is also suitable, first of all being determined and in particular quantified, in order then to set the threshold value as a function thereof.

In a suitable embodiment, several value ranges are defined for the noise value, each of which is assigned a threshold value. The range of values in which the noise value is located is then determined, and then that Threshold value selected and set, which is assigned to the determined value range. In this way, each noise value is assigned a sufficiently suitable threshold value in a simple manner, so that overall there is an assignment, for example in the form of a table, from which the most suitable threshold value in a particular situation is selected and then set. This is based on the consideration that the noise value lies within a certain range of values, which is now advantageously divided into several, in particular, interrelated intervals in order to implement a setting of the threshold value that is dependent on the noise value.

For example, the noise value is a level of the noise in the vicinity of the hearing aid. The level is usually given in dB. The range of values then ranges, for example, from -90 to -40 dB and is divided into approximately 10 to 20 value ranges, each with, for example, 5 dB. A separate threshold value is then assigned to each value range. When the hearing aid is in operation, the level of the noise is then measured and the threshold value which is assigned to the range of values in which the measured level is located is then set. The level is measured, for example, using a noise estimator, i.e. a so-called "noise estimator", e.g. based on a "minimum statistics" approach.

The assignment of threshold values to the value ranges takes place, for example, as part of a fitting session with the acoustician or by the hearing device wearer himself, for example as part of a calibration process. It is particularly important here that defined noise values are available or at least can be measured reliably. The assignment can be made by means of a pure calibration measurement and then available as a table and stored on the hearing device, or the assignment is made by a functional assignment, which is, for example, an approximation of the result of the calibration measurement. In the latter variant, for example, an upper and a lower limit are assumed for the threshold value, in particular an upper limit for low levels, for example below -75 dB, and a lower limit for high levels, for example above -60 dB, and linear extrapolation between them. This then needs to be advantageous only a suitable upper and a lower limit to be determined, as well as those value ranges over which extrapolation is then carried out.

In an expedient embodiment, the threshold value is recalibrated repeatedly during normal operation of the hearing device, in particular as described above with regard to the user-dependent determination of the optimal threshold value. As a result, the user-dependent threshold value is in particular continuously calibrated and, over time, is adapted better and better to the current hearing aid wearer. This corresponds in particular to a training session for the hearing aid, which is expediently ended after a certain training period. The user-dependent threshold value is then in particular fixed.

The hearing aid according to the invention has a natural voice recognition which is designed to carry out the method in one of the above-mentioned configurations. Depending on the result of the self-voice recognition, the hearing aid is then switched to a suitable operating mode for the respective situation. In one variant, switching also takes place using the self-voice recognition.

Fig. 1

ein Hörgerät mit einer Eigenstimmenerkennung,

Fig. 2

eine graphische Darstellung der Ergebnisse einer Messung zur Erkennung der eigenen Stimme eines Hörgeräteträgers,

Fig. 3

eine graphische Darstellung der Ergebnisse einer anderen Messung zur Erkennung der eigenen Stimme eines Hörgeräteträgers.

An exemplary embodiment is explained in more detail below with reference to a drawing. In it show:

Fig. 1

a hearing aid with natural voice recognition,

Fig. 2

a graphical representation of the results of a measurement for recognizing the own voice of a hearing device wearer,

Fig. 3

a graphical representation of the results of another measurement for recognizing the own voice of a hearing aid wearer.

In Fig. 1 a hearing aid 2 is shown schematically. This is designed as a so-called BTE device and is worn behind the ear by a user. In one variant, the hearing aid 2 is an ITE device and is worn in the ear. Also other types of hearing aids are generally suitable. The hearing aid 2 has a microphone 4 for recording noises from the surroundings of the hearing aid 2. A recorded sound is processed as a signal in a control unit 6 of the hearing aid 2 and processed for output via a loudspeaker 8. Usually the signal, ie the noise, is amplified.

The hearing aid furthermore has a natural voice recognition 10, which in the exemplary embodiment shown is part of the control unit 6. The control unit 6, the natural voice recognition 10, the microphones 4 and the loudspeaker 8 are suitably connected to one another. Furthermore, the hearing aid 2 can be operated in various operating modes, between which switching is carried out by means of the control unit 6 or the self-tuning detection 10. The self-voice recognition 10 analyzes the recorded noises and assigns them to certain types of noise G1, G2, for example the type of noise G1 "own voice" or the type of noise G2 "foreign voice". Depending on the recognized type of noise G1, G2, a switch is then made to a suitable operating mode. For the recognition, the natural voice recognition 10 generates a feature value M and compares it with a threshold value S in order to decide which type of noise G1, G2 the analyzed noise is. This is related to the Fig. 2 and 3 described in more detail below.

The Fig. 2 and 3 each show results of a measurement in which a sound was recorded and analyzed several times in succession. Two different types of noise G1, G2 were used, namely on the one hand the hearing aid wearer's own voice and on the other hand a foreign voice. The self-voice recognition 10 of the hearing device 2 first analyzes the recorded noise with the aim of assigning it a feature value M, which provides information as to whether the noise is of one or the other noise type G1, G2. In the present case, this was realized by a pair of filters with two filters that have different filter profiles. The filters are designed in such a way that one filter attenuates its own voice as much as possible and the other filter the foreign voice. By comparing the A characteristic value M is generated in the case of two different attenuations for the same noise.

The large number of characteristic values M which were recorded in the course of the measurements are shown in the Fig. 2 and 3 shown and plotted against a noise value R, here the level of noise in the environment. The noise figure is given here in decibels (dB). The noise value R is measured, for example, using an interference noise estimator. The feature values M are also each assigned to one of two groups, depending on which type of noise G1, G2 has actually been presented to the hearing aid. The characteristic values M, which were generated in the analysis of one's own voice as noise type G1, are shown in light gray, and the characteristic values M, which were generated in the analysis of the foreign voice as noise type G2, are shown in black. The measurements of the Fig. 2 and 3 now differ in that these results show for different hearing aid wearers, ie at least their own voice is different.

Is clearly recognizable in the Fig. 2 and 3 that when a foreign voice is present, a smaller characteristic value M is predominantly generated than when the own voice is present. This makes it possible to define a threshold value S with which a specifically generated characteristic value M is compared in order to decide which type of noise G1, G2 is present. In the exemplary embodiment, a sound is recognized by the self-voice recognition 10 as its own voice if the feature value M is greater than the threshold value S, and as a foreign voice if the feature value M is less than the threshold value S.

Conventionally, only a fixed threshold value S is used in order to be compared with the feature value M in any situation and environment. Like from the Fig. 2 and 3 it becomes clear that this may be insufficient. Rather, it can be seen that the use of different threshold values S makes sense in different environments. A first environmental dependency is that the generation of the feature value M is strongly dependent on the noise value R. For low noise values R are for your own voice still comparatively large characteristic values M are generated, but with a larger noise value R the difference to the characteristic values M of the foreign voice is significantly smaller. Therefore, a lower threshold value S is advantageously chosen for larger noise values R.

In Fig. 2 the optimum threshold values S are entered for individual value ranges W of the noise value R, namely as gray horizontal bars. As a result, a threshold value S is effectively assigned to a specific value range W, so that there is an overall assignment Z1 in the manner of a table. The hearing aid 2 then determines, on the one hand, a characteristic value M for a sound that has just been recorded and also the surroundings, in this case the noise value R, ie effectively the level or volume of the noise that is superimposed on the noise. Before the comparison with the characteristic value M, the threshold value S is then set depending on the environment, namely to the threshold value S which is assigned to the value range W in which the determined noise value R lies. Characteristic value M is thereby compared with a threshold value S adapted in the given situation and an optimal result is achieved in distinguishing between one's own voice and that of someone else.

Instead of the table-like assignment Z1 of the optimal threshold values S to the value ranges W, a simplified assignment Z2 is alternatively used. Such is also in Fig. 2 shown, namely as a dark gray, step-like line. It is assumed in a simplified manner that a maximum threshold value Smax is sufficient below a low noise value Rmin and a minimum threshold value Smin above a high noise value Rmax. In between there is an extrapolation of the threshold values S, here according to a linear relationship with respect to the selected representation. Overall, the simplified assignment Z2 virtually smoothes the assignment Z1 with the optimal threshold values S. The assignment Z2 is saved in a variant as a simple table, alternatively a function for the calculation is saved.

When comparing the Fig. 2 and 3 a further dependence on the environment of the characteristic values M becomes clear, namely the person of the hearing device wearer. In Fig. 3 are on the one hand as well as in Fig. 2 an assignment Z1 of optimal threshold values S to certain value ranges W is shown as a gray horizontal bar. In addition, the same simplified assignment Z2 is from Fig. 2 in the Fig. 3 registered, namely again as a dark gray, stair-like line. When comparing the simplified assignment Z2, which for the hearing aid wearer Fig. 2 was determined with the optimal threshold values S for the other hearing aid wearer Fig. 3 according to the assignment Z1 it is immediately clear that the in Fig. 2 determined assignment Z2 in Fig. 3 is not optimal. Therefore, the threshold value S is advantageously also set as a function of the user, ie as a function of the person of the hearing aid wearer.

Overall, the threshold value S is accordingly preferably set in two ways depending on the environment, namely on the one hand depending on the user and on the other hand depending on the noise value R measured at a given point in time. Which threshold value S is then set specifically, i.e. one or at the assignments Z1, Z2, i.e. Which threshold values S are available for selection is expediently determined in a calibration process. This is carried out either as part of a fitting session at the acoustician, by the hearing aid wearer himself, automatically by the hearing aid as part of an online optimization, or a combination thereof.

In order to determine an optimal threshold value S for a given hearing aid wearer and for a specific noise value R, there are in particular the above in connection with the Fig. 2 and 3 described measurements. In this case, noises of a known noise type G1, G2 are analyzed and the characteristic values M determined in this way are used as typical characteristic values M in order to determine a suitable threshold value S. When using two different noise types G1, G2, two different statistical distributions of feature values M are then determined, for example, and then a threshold value S is selected in between. However, the use of only one noise type G1, G2 is also conceivable. The calibration is carried out in one variant by use previously known noise types G1, G2, so that the correct assignment is trained. In another variant, the calibration takes place during normal operation of the hearing device 2 by generating feature values M in limited periods of time from a few seconds to a few tens of seconds and on the assumption that the extrema of the feature values M determined in a given period of time are determined with a certainty with sufficient certainty Allocate sound types G1, G2. For example, it is assumed that the generation of a maximum characteristic value M was caused by one's own voice and the generation of a minimum characteristic value M by an external voice. These extremes are then used to determine an optimal threshold value S, which can be and is expediently further adapted by continuous calibration in the further operation of the hearing device 2.

LIST OF REFERENCE NUMBERS

2

hearing Aid

4

microphone

6

control unit

8th

speaker

10

Own voice recognition

G1, G2

type of noise

M

feature value

R

noise value

Rmin

low noise figure

Rmax

high noise figure

S

threshold

Smin

minimum threshold

Smax

maximum threshold

W

value range

Z1, Z2

assignment

Claims (11)

Method for operating a hearing aid (2), - wherein a noise is recorded by means of a microphone (4), the noise is analyzed with regard to its correspondence with the hearing aid wearer's own voice and a characteristic value (M) is generated which indicates how closely the noise corresponds to the hearing aid wearer's own voice, - where your own voice is a type of noise (G1), the characteristic value (M) is compared with a threshold value (S), - The sound depending on whether the feature value (M) is above or below the threshold value (S) is recognized as a separate voice, and - The hearing aid (2) is switched between several operating modes depending on whether the sound was recognized as a separate voice, characterized,
that the threshold value (S) is determined as a function of the user and is set as an individual threshold value (S) by determining the threshold value (S) by means of a calibration method, in which the hearing aid wearer's own voice is recorded and several individual characteristic values (M) are generated, and at which finally the individual threshold value (S) is set depending on the individual characteristic values (M) generated. Method according to the preceding claim,
characterized,
that the threshold value (S) is calibrated by determining a maximum and a minimum feature value (M) over a limited period of time and the threshold value (S) being set between the minimum and the maximum feature value (M). Method according to one of the preceding claims,
characterized,
that the threshold value (S) is recalibrated repeatedly in normal operation when the hearing device wearer uses the hearing device (2). Method according to one of the preceding claims,
characterized,
that in addition to matching your own voice, the noise is also analyzed for a match with at least one other type of sound (G2). Method according to the preceding claim,
characterized,
that the other type of noise (G2) is a foreign voice, which is arranged frontally with respect to the hearing aid wearer. Method according to the preceding claim,
characterized,
that a different type of noise (G2), in particular a foreign voice, is recorded in the calibration process before or after the recording of one's own voice, and that several characteristic values (M) are also generated here, depending on which the threshold value (S) is set. Method according to one of the preceding claims,
characterized,
that the feature value is generated by means of a filter pair, one of the filters being configured for maximum attenuation of one's own voice and the other filter configured for maximum attenuation of a foreign voice. Method according to one of the preceding claims,
characterized,
that the threshold value (S) is set depending on the environment by determining a noise value (R) and the threshold value (S) is set depending on the noise value (R). Method according to the preceding claim,
characterized,
that several value ranges (W) are defined for the noise value (R), each of which is assigned a threshold value (S), that value range (W) is determined in which the noise value (R) lies, and that threshold value (S) is selected and set which is assigned to the determined value range (W). Method according to one of the two preceding claims and
one of claims 7 or 8,
characterized,
that the threshold value (S) is calibrated in normal operation by repeatedly determining the noise value (R) and calibrating the threshold value (S) depending on it. Hearing aid (2) with a microphone (4) for recording a noise, and with a natural voice recognition (10), which is designed such that the noise is analyzed with regard to its agreement with the hearing aid wearer's own voice and a characteristic value (M) is generated which indicates how closely the noise corresponds to the hearing aid wearer's own voice, - where your own voice is a type of noise (G1), the characteristic value (M) is compared with a threshold value (S), the sound is recognized as a separate voice depending on whether the feature value (M) is above or below the threshold value (S), - depending on whether the sound was recognized as a separate voice, it is switched between several operating modes, - The threshold value (S) is determined as a function of the user and is set as an individual threshold value (S) by determining the threshold value (S) by means of a calibration method, in which the hearing aid wearer's own voice is recorded and several individual characteristic values (M) are generated, and at which finally the individual threshold value (S) is set depending on the individual characteristic values (M) generated.

Images