Hearing devices and methods for operating hearing devices

Abstract

A hearing device (2) and a method for operating the hearing device are provided, the hearing device having at least one microphone (4) configured to receive sound signals (6) and convert them into input signals (8) within a global frequency range (Fg), the hearing device having a signal processing device (10) for processing the input signal (8) in a lower frequency range (Fu), the lower frequency range being a part of the global frequency range (Fg), the hearing device having a detector (12) for identifying interfering noise (14) having frequency components within the lower frequency range (Fu) and outside the lower frequency range, i.e., in a higher frequency range (Fo) above the lower frequency range (Fu), the hearing device being configured such that the detector identifies the interfering noise based on the frequency components of the interfering noise in the higher frequency range (Fo).

Description

Technical Field

[0001] This invention relates to a hearing device and a method for operating the hearing device. Background Technology

[0002] Hearing devices are generally used to output sound signals to users of hearing devices. Especially when hearing devices are used to supply users with hearing impairments, the hearing devices are constructed to receive sound signals from the environment, process them, and finally modify (i.e., amplify) them in a way that at least partially compensates for the hearing impairment before outputting them.

[0003] In a given environment, there are typically both useful noise and interfering noise. The former should be reproduced in a way that is as easy for the user to understand as possible, while the latter should be suppressed in the most comprehensive way possible, thereby improving the operation of the hearing equipment and its use by the user. However, effective suppression of interfering noise requires identifying these interfering noises as effectively as possible. In principle, there are different types of interfering noise, each with varying degrees of difficulty in identification. Interfering noise differs, for example, in its bandwidth, amplitude, and / or location within the total spectrum recorded by the hearing equipment. Examples of different types of interfering noise are thus: stable (i.e., narrow-band and with low time variability) interfering noise, unstable, here especially transient and / or impulsive (i.e., wideband with high time variability) interfering noise, loud (i.e., high-amplitude) interfering noise, soft (i.e., low-amplitude) interfering noise, high-frequency (dominated by high-frequency components) interfering noise, and low-frequency (dominated by low-frequency components) interfering noise. Examples of non-transient interfering noise are stable noises, such as fan noise or background noise, often referred to by the term "cocktail party."

[0004] When hearing devices are in operation, suppression of interference noise, especially transient and / or impulsive interference noise, depends on its reliable identification. Hearing devices typically operate within a specific frequency range, the so-called operating range, within which the input signal is modified to suit the user, preferably for compensating for the user's hearing impairment. To facilitate this modification of the input signal using signal processing devices, it is particularly preferable that the hearing device stores the user's audiogram, which is confined to the operating range. Outside the operating range, the input signal is not modified, at least not actively modified for compensating the user's hearing impairment. The operating range is typically only a subrange of the acoustic spectrum. This acoustic spectrum is specifically defined as the frequency range from 20 Hz to 20 kHz. This limitation to a subrange of the acoustic spectrum is usually due to the technical limitations of the hearing device, as it typically cannot process arbitrarily high frequencies.

[0005] Interference noise is identified within the operating range, for example, by identifying particularly loud frequency components, i.e., frequency components with a predetermined minimum amplitude, as noise. For instance, interference noise is identified by identifying frequencies exceeding the minimum amplitude within one or more frequency ranges. To suppress interference noise, the gain of the hearing device is reduced, for example, within these frequency ranges, so that the interference noise is amplified less compared to the useful noise in other frequency ranges.

[0006] However, the problem lies in identifying interfering noise with relatively small amplitudes—that is, interfering noise with amplitudes similar to or smaller than the possible or common amplitudes of useful noise. This interfering noise is also referred to as "soft interfering noise," "less loud interfering noise," or "gentle interfering noise." This soft interfering noise cannot be, or can only be poorly, identified by the detector, which needs to exceed a pre-given minimum amplitude to identify it. That is, the minimum amplitude should be chosen so as not to affect useful noise, especially speech, as little as possible. Therefore, speech often also has transient and / or impulsive components; however, speech is useful noise, not interfering noise, and should therefore be preserved. Thus, the detector can ultimately only identify interfering noise starting at a specific volume; that is, the minimum amplitude cannot be chosen arbitrarily small. Therefore, the detector can only poorly identify interfering noise with relatively small amplitudes, especially interfering noise with amplitudes within the common amplitude range of speech. Correspondingly, in suppressing interfering noise, a conflict arises between minimizing the impact of incorrect identification on useful noise, especially speech, on the one hand, and suppressing soft interfering noise on the other.

[0007] Different hearing devices are described in US 2013 / 0 156 208 A1, EP 3 793 209 A1 and DE 10 2015 204 253 A1. Summary of the Invention

[0008] In this context, the technical problem to be solved by the present invention is to improve the identification of interference noise during the operation of a hearing device. In particular, it is to improve the identification of soft-sounding interference noise. To this end, an improved hearing device and an improved method for operating the hearing device are provided.

[0009] According to the present invention, the aforementioned technical problems are solved by a hearing device having the features of the present invention and by a method having the features of the present invention. Advantageous designs, extensions, and modifications are the subject of the following description. The description of the hearing device also applies to the method. If the steps of the method are given below, it is preferred that the design for the hearing device is obtained in such a way that the hearing device is constructed, in particular by means of a control unit that is part of the hearing device, to perform one or more of these steps.

[0010] The core idea of ​​this invention is specifically to identify interference noise within a frequency range in which, when designing a detector for interference noise, target conflicts are resolved by ensuring that useful noise is absent or present only to a small degree. Thus, compared to frequency ranges where useful noise is expected to be present, a smaller minimum amplitude can be advantageously used for identification without significantly increasing the risk of false identification. In this way, even faint interference noise is better identified, at least faint interference noise having frequency components within the stated frequency range is also better identified. This is particularly true for broadband interference noise, especially transient and / or impulsive interference noise, which typically has frequency components outside the operating range of the hearing device (particularly as described above) and / or in a higher frequency range (or upper frequency range) outside the speech frequency range.

[0011] Therefore, the hearing device according to the invention has at least one microphone configured to receive sound signals and convert them into input signals across the entire frequency range. Furthermore, the hearing device has a signal processing means for processing the input signal within a lower frequency range (or lower frequency range) that is part of the overall frequency range. Thus, the hearing device has an operating range (particularly as described above) that at least and preferably only includes the lower frequency range. Furthermore, the hearing device has a detector for identifying interference noise having frequency components within the lower frequency range and outside the lower frequency range, i.e., within a higher frequency range above the lower frequency range. Preferably, the higher frequency range is directly adjacent to the lower frequency range, but this is not mandatory. Moreover, the hearing device is configured such that the detector identifies the interference noise based on the frequency components of the interference noise within the higher frequency range. This, in particular, realizes the core idea described above. In general, the hearing device is therefore configured to identify interference noise during operation.

[0012] For simplicity, the term "identifying interference noise" will be used here. However, more accurately, "identifying interference noise" should be understood as identifying interference signals in the input signal, which are generated by acoustic interference noise that is part of the ambient sound signal. The result is thus the identification of interference noise in the environment.

[0013] The higher and lower frequency ranges should not overlap. The higher frequency range is preferably not part of the operating range; however, this is not mandatory, and designs in which the operating range also includes, wholly or partially, the higher frequency range are possible and suitable. In particular, and more importantly, the higher frequency range should be characterized in such a way that less useful noise is present and / or expected to be present in the higher frequency range compared to the lower frequency range. In other words, the SNR (signal-to-noise ratio) is greater in the higher frequency range, especially for the aforementioned subtle interfering noise, than in the lower frequency range (here, when considering SNR, the interfering noise considered is the "signal," while all other noise and interfering noise constitute "noise"). Outside the lower frequency range, the speech component is generally also particularly small, effectively reducing false recognition.

[0014] This invention is based in particular on the consideration that, generally speaking, the sound signal, and specifically the interference noise, is not limited to the operating range of the hearing device, but may also be located outside the operating range, though still within the aforementioned acoustic spectrum (defined from 20 Hz to 20 kHz). In particular, transient and / or pulse-like interference noise is broadband, thus spanning a wide frequency range, that is, it also has high-frequency components that are typically located outside the operating range (the terms "component" and "frequency component" are generally considered equivalent and can be used interchangeably). Interference noise is considered broadband, especially when the relevant frequencies span at least one octave (or octave). The identification of such interference noise is less reliable, particularly at low frequencies, especially within the operating range of the hearing device, where other stable or unstable interference noise is often also present, often louder and with higher amplitude. Additionally, unstable components of speech can also lead to misidentification, that is, misidentifying speech that is actually useful noise as interfering noise.

[0015] However, it has been observed that the amplitude of high-frequency interference noise (i.e., interference noise with high-frequency components) is often smaller than that of low-frequency interference noise, i.e., interference noise that is generally located within the operating range. For example, in a restaurant or cafeteria, the clinking noise (also known as "clinking") produced when cutlery and utensils collide with each other is a soft, transient, and / or pulse-like interference noise, which also has a high-frequency component that is generally located outside the operating range. Now, this interference noise can be identified better, especially with the aid of the hearing device according to the invention, because the soft interference noise is also identified in the higher frequency range. Here, the risk of false identification is significantly reduced because useful noise is expected to be absent or only present in small amounts in the higher frequency range, since useful noise is mainly or only located in the lower frequency range.

[0016] Therefore, in particular, one advantage of the present invention is the better identification of faint transient and / or pulsed interference noise. This is achieved, especially by monitoring the higher frequency range for identification of such interference noise. Thus, transient and / or pulsed interference noise is identified based on its frequency components outside the lower frequency range, thereby avoiding the problems described when identifying it in that lower frequency range. Here, the fact that interference noise is broadband and has frequency components in both the lower and higher frequency ranges is utilized.

[0017] Preferredly, signal processing devices are used to suppress interference noise only in the lower frequency range, and not necessarily in the higher frequency range, when necessary. Here, it is assumed that the interference noise identified in the higher frequency range is correspondingly broadband and extends into the lower frequency range. When identifying interference noise in the higher frequency range, it is meaningful to suppress which frequency components in the lower frequency range. This is done, for example, by determining experimentally beforehand, or by estimating based on the common bandwidth of interference noise that also has components in the higher frequency range. A suitable design is to simply suppress interference noise within a predefined frequency range in the lower frequency range, without actually knowing whether it includes interference noise. This utilizes the knowledge that interference noise has a wide spectrum and exists only for a short time, and is therefore very likely located within a predefined frequency range. The predefined frequency range is, for example, from the upper limit of the lower frequency range to 1 kHz to 5 kHz within the lower frequency range. A suitable design is also to suppress interference noise by averaging the input signal over a specific time period, for example, 1 to 5 seconds, thereby smoothing it. However, the details of suppressing interference noise are not important here; instead, the focus is on identifying interference noise as reliably as possible.

[0018] The following design is also advantageous, in which, in addition to the already mentioned detector (the first detector), a second detector, such as a pulse detector, monitors a lower frequency range. This second detector is then triggered as an additional criterion for suppression to ensure the actual presence of the corresponding interference noise. For example, suppression occurs only when both the first detector and the second detector identify transient and / or pulse-like interference noise in the higher frequency range, respectively. For these detectors, it is preferable to divide the lower frequency range more coarsely than for signal processing, for example, dividing it into only two frequency bands, such as a first band of 100Hz to 1000Hz and a second band of 1000Hz to 12kHz.

[0019] Hearing devices are preferably designed for users with hearing impairments, i.e., users with hearing defects. For this purpose, the hearing device has a microphone, as mentioned above, which receives sound signals from the environment and generates an electrical input signal. This electrical input signal is fed to a signal processing device, also mentioned, for processing. The signal processing device is preferably part of the control unit of the hearing device. Specifically, the processing is performed based on the individual user's audiogram associated with the hearing device, thereby compensating for the individual user's hearing defect. Preferred processing corresponds to amplification. The signal processing device outputs an electrical output signal as a result, which is then, for example, again output as an audio signal through the earpiece of the hearing device to the user, generated by the earpiece based on the output signal.

[0020] Alternatively, the hearing device is simply a pair of headphones, preferably equipped with interference noise suppression for targeted suppression of interference noise, which in turn benefits from improved recognition of interference noise.

[0021] The specific design of the detector is initially irrelevant; what matters is that the detector operates in the higher frequency range and searches for interfering noise there. In principle, a conventional detector would be suitable—one that would otherwise be used in lower frequency ranges—but is now used instead to monitor the higher frequency range. This can be achieved, for example, by limiting the detector to the higher frequency range, either by ensuring the detector only monitors the higher frequency range or by feeding only the higher frequency range to the detector.

[0022] In a suitable design, the hearing device has a filter bank that divides the input signal into multiple channels, assigning a frequency band to each channel. The filter bank is either integrated into the signal processing unit or constructed separately. The filter bank has multiple channels, particularly at least three, but typically a double-digit number. The first number of channels then forms a lower frequency range, while a second number similarly forms a higher frequency range, which is then fed to the detector. In any case, the filter bank also facilitates the suppression of interfering noise because the gain can be selectively adjusted in each individual channel, thereby precisely reducing those components belonging to interfering noise. A design is also possible in which the filter bank is confined to the lower frequency range, while the higher frequency range is directed to the detector alongside the filter.

[0023] Suitablely, the detector is a pulse detector, preferably a gradient-based pulse detector. For a gradient-based pulse detector, the key feature is that it detects transient and / or pulse-like interference noise based on a specific increase in amplitude over time.

[0024] In a suitable design, the detector is configured as a horizontal detector, whereby it identifies interfering noise as interfering noise if the amplitude in a higher frequency range exceeds a minimum amplitude. The minimum amplitude depends on the user's personal preference and hearing impairment, and can be appropriately adjusted by the user. When the minimum amplitude is exceeded, the detector reaches a higher frequency range, thereby identifying the interfering noise. Specifically, the minimum amplitude is selected to be lower than (i.e., half the minimum amplitude) selected for a detector monitoring a lower frequency range.

[0025] Alternatively or additionally, the detector is suitably constructed as a gradient detector, whereby the detector identifies the interfering noise as interfering noise if the amplitude of the interfering noise in the higher frequency range undergoes minimal change during a predefined time interval. Therefore, the gradient detector monitors how much the amplitude changes (e.g., increases) within the predefined time interval, and triggers the detector only if the amplitude undergoes at least a minimal change (i.e., a minimum amplitude change) within that predefined time interval. This predefined time interval depends in particular on the sampling rate of the input signal in the hearing device and is, for example, in the range of 1 ms to 20 ms. The minimum change, like the minimum amplitude, is individual, and the description above applies similarly.

[0026] The following design is also suitable, in which, alternatively or additionally, the detector is constructed as a wavelet detector, thereby identifying the interfering noise as interfering noise based on the spectral shape of the interfering signal. The spectral shape specifically refers to the frequency-dependent amplitude trend of the interfering noise. To identify the interfering noise, a wavelet with an ideal spectral shape is pre-defined and then compared with the actual interfering noise, particularly by correlating the wavelet with the noise. The stronger the correlation between the wavelet and the interfering noise, the more closely the interfering noise corresponds to the wavelet. This is essentially analogous to image comparison. In this way, specific interfering noise with a substantially known spectral shape can be targeted for searching and identification. For example, a simple pulse can be used as the wavelet.

[0027] Here, the additional use of a second detector for the lower frequency range is not mandatory, but it serves as a complement to the first detector described herein for the higher frequency range. For example, it is advantageous for identifying narrowband interference noise in the lower frequency range, which is particularly impossible using the detector described herein. Correspondingly, in one design, the second detector for the lower frequency range, acting as a horizontal detector, has a higher minimum amplitude than the detector for the higher frequency range. Due to the problem described at the beginning, to identify interference noise in the lower frequency range, a minimum amplitude greater than the common or maximum achievable amplitude of useful noise in the lower frequency range is required. Because useful noise is absent or present only sparingly in the higher frequency range, a correspondingly smaller minimum amplitude is possible and advantageous, as it also better identifies faint interference noise. In other respects, the description of the first detector also applies similarly to the optional second detector.

[0028] As already noted, the interference noise that can be identified using a detector is preferably transient and / or pulsed interference noise. This interference noise is broadband, and therefore typically has components in the higher frequency range (and generally also has components in the lower frequency range), meaning that it can be identified using a detector there. At the same time, this interference noise is often so quiet that it may not be reliably identifiable in the lower frequency range.

[0029] In a favorable design, the interference noise is or will be generated by the clinking of cutlery and / or knives and forks. This generation mechanism yields a specific spectrum for the interference noise; that is, observed in the frequency domain, the interference noise has a characteristic distribution (charakteristisches profile), namely, a broad spectrum and low amplitude, because the interference noise is a soft, transient, and / or impulsive interference noise. The characteristic distribution in the frequency domain has already been mentioned above in conjunction with wavelet detectors and is referred to there as the spectral shape; the two terms are equivalent.

[0030] However, the specific distribution (i.e., spectral shape) of the interfering noise is secondary; more important is the impulsiveness of the interfering noise. This interfering noise is often very loud but difficult to detect because other transient and / or impulsive interfering noises also appear within the frequency range of the hearing device (e.g., 100 Hz to 10 kHz). These are not interfering noises but useful noises (e.g., plosives in speech) and therefore should not be suppressed. Therefore, it is crucial to first distinguish between desirable and undesirable transient and / or impulsive interfering noises—that is, useful transient and / or impulsive noises on the one hand, and interfering noises on the other.

[0031] As mentioned above, the higher frequency range is preferably not part of the operating range of the hearing device. Therefore, in a suitable design, the lower frequency range is the operating range of the signal processing device, thus limiting the processing of the input signal by the signal processing device to the lower frequency range. Specifically, for this purpose, the operating range and the lower frequency range are the same. The preferred processing is amplification. Therefore, the signal processing device does not process, and preferably does not amplify, the higher frequency range; it is only used for other functions of the detector and, possibly, the hearing device, which are further unrelated here. In particular, amplification intended to compensate for the user's hearing loss is not performed in the higher frequency range, but only within the operating range.

[0032] Alternatively or additionally, the speech frequency range is wholly or at least primarily (i.e., particularly up to at least 90%) encompassed by a lower frequency range. In particular, the speech frequency range accurately defines the frequency range in which the speech occupies. The speech frequency range extends, for example, from 100 Hz to 4 kHz, to 6 kHz, to 8 kHz, or to 12 kHz, especially when overtones are considered together. At that time, a higher upper limit applies. However, the precise definition of the speech frequency range is not important at first; what is more important is that speech is mainly or completely included in the lower frequency range, and correspondingly absent or negligible in the higher frequency range.

[0033] Ideally, the speech frequency range should be completely encompassed by the operating range. Designs where the operating range and speech frequency range have the same upper and / or lower limits are also possible and suitable. However, designs where the operating range is larger or smaller than the speech frequency range are also conceivable. Especially in designs where the operating range also includes a higher frequency range, the operating range is larger than the speech frequency range, and the speech frequency range is preferably limited to a lower frequency range or even lower.

[0034] The higher frequency range appropriately has an upper limit, which corresponds at least to the upper limit of the previously mentioned speech frequency range. This ensures that speech in the higher frequency range is not mistakenly identified as transient and / or impulsive interference noise.

[0035] Preferably, the lower frequency range has an upper limit of 12kHz, 8kHz, 6kHz, or 4kHz. Simultaneously, this upper limit is preferably, but not necessarily, a lower limit of the higher frequency range. The 12kHz upper limit is based on the consideration that speech is thus entirely within the lower frequency range, thereby avoiding false identification of interfering noise by the detector in the higher frequency range. Furthermore, for the same reason, typical hearing devices usually have an operating range with an upper limit of 12kHz. The lower frequency range also has a lower limit, the exact value of which is irrelevant here. For example, a lower limit of 20Hz is suitable.

[0036] The higher frequency range has an upper limit, which is appropriately defined by the technical boundary conditions of the hearing device, particularly by the finite frequency range of the filter bank and / or the finite sampling rate of the A / D converter (i.e., analog-to-digital converter) that digitizes the input signal from the microphone. In a preferred design, the higher frequency range has a lower limit of 12 kHz and / or an upper limit of 16 kHz. This lower limit is derived, in particular, based on the same considerations as the upper limit of the lower frequency range. However, the exact value of the upper limit is, in principle, irrelevant and, in a suitable design, is very generally derived by the technical boundary conditions of the hearing device. The 16 kHz upper limit mentioned here is derived, in particular, by the common technical boundary conditions of hearing devices, especially by the 32 kHz sampling rate of the A / D converter in the hearing device.

[0037] In a suitable design, the hearing device has an A / D converter for digitizing the input signal, and the higher frequency range has an upper limit that corresponds to a maximum of half the sampling rate of the A / D converter. Therefore, the higher frequency range is derived through pre-defined technical boundary conditions of the A / D converter. Depending on the design of the hearing device, a corresponding upper limit is thus derived for the higher frequency range. In a suitable design, the A / D converter has a sampling rate of 32 kHz, so the upper limit corresponds to a maximum of 16 kHz. Here, the sampling rate specifically refers to the actual set sampling rate of the A / D converter, i.e., the actual set sampling rate. For example, the A / D converter itself may have a significantly higher possible sampling rate, such as several megahertz, but as an oversampled A / D converter (i.e., as an "oversampled A / D-converter"), it operates at a significantly lower actual set sampling rate of 32 kHz. Furthermore, the sampling rate, especially the actual sampling rate set in an oversampling A / D converter, is typically affected by tolerances. These tolerances are primarily due to the deviation of the oscillation frequency of the quartz crystal used for timing the A / D converter from the ideal frequency, and also due to the oscillation frequency being divided by multiples of 2 up to the actual sampling rate set. In other words, an actual sampling rate of 32kHz actually corresponds to a sampling rate of 32kHz plus or minus, for example, a 5% tolerance.

[0038] The method according to the invention is used to operate a hearing device, particularly a hearing device as described above. The hearing device includes a microphone, a signal processing unit, and a detector. In the method, i.e., while the hearing device is operating, a sound signal is received by means of the microphone across the entire frequency range and converted into an input signal. Furthermore, the signal processing unit processes only, or at least, the input signal in the lower frequency range, which is part of the overall frequency range. Then, the detector identifies interfering noise, which has frequency components both within the lower frequency range and outside the lower frequency range, i.e., in the higher frequency range above the lower frequency range, wherein the detector identifies the interfering noise based on its frequency components in the higher frequency range. Advantageous and preferred extensions are derived from what has already been described. The interfering noise in the input signal is suitably suppressed, and the input signal is output via the earpiece. Attached Figure Description

[0039] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings. Wherein:

[0040] Figure 1 Accordingly, hearing devices are schematically shown.

[0041] Figure 2 The spectrum is schematically shown accordingly.

[0042] Figure 3 Correspondingly, another spectrum is schematically shown.

[0043] Figure 4 Accordingly, schematically shown Figure 3 A portion of the spectrum. Detailed Implementation

[0044] exist Figure 1 An embodiment of a hearing device 2 according to the present invention is shown. The hearing device 2 has at least one microphone 4, which is configured to receive a sound signal 6 within a general frequency range Fg and convert it into an input signal 8. Furthermore, the hearing device 2 has a signal processing unit 10 for processing the input signal 8 within a lower frequency range (or lower frequency range) Fu, which is part of the general frequency range Fg. Therefore, the hearing device 2 has an operating range 4, the operating range Fa of which at least, and in the embodiment shown herein, even only includes the lower frequency range Fu. Furthermore, the hearing device 2 has a detector 12 for identifying interference noise 14 having frequency components within the lower frequency range Fu and outside the lower frequency range Fu, i.e., within a higher frequency range Fo above the lower frequency range Fu. This is in Figure 2 As shown in the figure, Figure 2 An exemplary spectrum of the input signal 8 is shown, i.e., its amplitude A as a function of frequency F. Here, for clarity, interference noise 14 and useful noise 16 are shown separately; however, it is evident that interference noise 14 and useful noise 16 are mixed into the input signal 8, and may contain other useful noise and / or interference noise, which are not explicitly shown for clarity. Furthermore, the hearing device 2 is configured such that the detector 12 identifies interference noise 14 based on its frequency components within the higher frequency range Fo.

[0045] Here, the higher frequency range Fo is directly adjacent to the lower frequency range Fu, but this is not mandatory. In the illustrated embodiment, the higher frequency range Fo and the lower frequency range Fu do not overlap. Furthermore, the higher frequency range Fo is not part of the operating range Fa; however, this is not mandatory in itself, and thus a design in which the operating range Fa also includes all or part of the higher frequency range Fo is possible. In particular, the higher frequency range Fo is characterized in such a way that less useful noise 16 exists and / or is expected to exist in the higher frequency range Fo compared to the lower frequency range Fu, as in... Figure 2 This can also be seen by example. Outside the lower frequency range Fu, the speech components (which are useful noise 16) are generally also particularly small.

[0046] Generally speaking, the sound signal 6, and specifically the interference noise 14, is not limited to the operating range Fa of the hearing device 2, but may also be located outside the operating range Fa of the hearing device 2. In particular, transient and / or pulse-like interference noise 14 is broadband, thus spanning a wide frequency range; that is, it also has high-frequency components that are typically located outside the operating range Fa, such as in... Figure 2 The interference noise 14 shown therein becomes clear (the terms "component" and "frequency component" are generally considered equivalent and can be used interchangeably). The identification of this interference noise 14 is less reliable, especially at low frequencies, for example, within the operating range Fa of the hearing device 2, where other stable or unstable interference noises are often also present (in...). Figure 2 (Not explicitly shown in the text) These interfering noises are often also louder, i.e., have a higher amplitude A. Additionally, unstable components of speech can also lead to misidentification, i.e., misidentifying speech that is actually useful noise 16 as interfering noise 14.

[0047] However, it has been observed that the amplitude A of high-frequency interference noise 14 is often smaller than the amplitude A of low-frequency interference noise 14 that is generally located within the operating range Fa. For example, in a restaurant or canteen, the clinking interference noise 14 produced when cutlery and utensils collide with each other is a soft, transient and / or pulse-like interference noise 14, which also has a high-frequency component that is usually located outside the operating range Fa, such as in Figure 2 As exemplarily illustrated in [the text]. Now, in particular, the hearing device 2 better identifies this interference noise 14 because the faint interference noise 14 is also identified in the higher frequency range Fo. Here, the risk of false identification is reduced because useful noise 16 is expected to be absent or only present in small amounts in the higher frequency range Fo, since useful noise 16 is mainly or only located in the lower frequency range Fu, as can also be exemplarily illustrated in [the text]. Figure 2 This is observed in the image. Therefore, here, in particular, the transient and / or pulse-like interference noise 14 is identified by monitoring the higher frequency range Fo. Thus, the transient and / or pulse-like interference noise 14 is identified based on the frequency components of the interference noise 14 outside the lower frequency range Fu. Here, it is taken advantage of the fact that the interference noise 14 is broadband and has frequency components within both the lower and higher frequency ranges Fu and Fa.

[0048] Suppression of interference noise 14 is performed, for example, by means of signal processing device 10, and thus, when necessary, only in the lower frequency range Fu, and not necessarily in the higher frequency range Fo.

[0049] The hearing device 2 shown here is intended for users with hearing impairments, i.e., users with hearing defects. For this purpose, the hearing device 2 has a microphone 4, as mentioned above, which receives sound signals 6 from the environment and generates an electrical input signal 8. The electrical input signal 8 is fed to a signal processing device 10, also mentioned, for processing (specifically, amplification). This signal processing device 10 is also part of the control unit 18 of the hearing device 2. Processing is performed based on the individual user's audiogram associated with the hearing device 2, thereby compensating for the individual user's hearing defect. The signal processing device 10 outputs an electrical output signal 20 as a result, which is then output to the user through the earpiece 22 of the hearing device 2.

[0050] In an alternative not shown, the hearing device 2 is simply an earphone, and then, for example, has interference noise suppression for targeted suppression of interference noise 14, which accordingly benefits from the identification of interference noise 14 as described herein.

[0051] The specific design of detector 12 is irrelevant here; what is more important is that detector 12 operates within the higher frequency range Fo and searches for interfering noise 14 there. In principle, a conventional detector 12 would be suitable, as it would otherwise be used for the lower frequency range Fu, but now it is used instead to monitor the higher frequency range Fo. This limitation to the higher frequency range Fo is achieved, for example, by either monitoring only the higher frequency range Fo or feeding only the higher frequency range Fo to detector 12.

[0052] In the exemplary design shown here, the hearing device 2 has a filter bank 24 that divides the input signal 8 into multiple channels, assigning a frequency band to each channel. The filter bank 24 is configured to be separate from the signal processing unit 10, although it can alternatively be integrated into the signal processing unit 10. The filter bank 24 has multiple channels, particularly at least three, but typically two-digit numbers. A first number of channels then form a lower frequency range Fu, while a second number of channels similarly form a higher frequency range Fo, which is then fed to the detector 2. In any case, the filter bank 24 also facilitates the suppression of interference noise 14, because the gain can be selectively adjusted in each individual channel, thereby specifically reducing those components belonging to the interference noise 14. A design is also possible in which the filter bank 24 is confined to the lower frequency range Fu, while the higher frequency range Fo is directed to the detector alongside the filter 14. Figure 1 This is the situation in China, according to Figure 1The input signal 8 is fed entirely to detector 12, which then controls signal processing device 10 to suppress interference noise. Alternatively or additionally, detector 12 (as already indicated) controls filter bank 24 for suppression. Furthermore, in a variant not shown, (as also already indicated) the input signal 8 is not fed entirely to detector 12, but only the higher frequency range Fo from filter bank 24 is fed to detector 12.

[0053] Here, detector 12 is exemplarily configured as a horizontal detector, whereby if the amplitude A of the interference noise 14 in the higher frequency range Fo exceeds the minimum amplitude M, detector 12 identifies the interference noise 14 itself. When the minimum amplitude M is exceeded, detector 12 is triggered in the higher frequency range Fo, thereby identifying the interference noise 14. Here, the minimum amplitude M is selected to be lower than (i.e., half the size) the minimum amplitude M selected for the detector monitoring the lower frequency range Fu. This detector for the lower frequency range Fu is not used here, but in an alternative not shown, it exists as a supplement to the detector 12 described herein for the higher frequency range Fo, for example, for identifying narrowband interference noise in the lower frequency range Fu, which is not possible using the detector described herein. Correspondingly, the detector for the lower frequency range Fu then has a higher minimum amplitude M than the detector 12 for the higher frequency range Fo. However, in order to identify the interfering noise 14 in the lower frequency range Fu, a minimum amplitude M is required that is greater than the common or maximum achievable amplitude A of the useful noise 16 in the lower frequency range Fu, such as from Figure 2 As can be seen in the image. Because there is no or only a small amount of useful noise 16 in the higher frequency range Fo, a correspondingly smaller minimum amplitude M is possible, and subtle interfering noise 14 is also identified.

[0054] Detector 12 is, for example, a gradient-based impulse detector. Design schemes as gradient detectors or wavelet detectors are also possible, as are combinations of the designs mentioned above.

[0055] As already mentioned, in the illustrated design, the higher frequency range Fo is not part of the operating range Fa of the hearing device 2. Therefore, the lower frequency range Fu is the operating range Fa of the signal processing device 10, thus limiting the processing of the input signal 8 by the signal processing device 10 to the lower frequency range Fu. Therefore, the signal processing device 10 does not process the higher frequency range Fo; it is only used for detector 12 and possibly other functions of the hearing device 2 that are further irrelevant here. Here, it is irrelevant whether filter bank 24 only transmits the lower frequency range Fu to signal processing device 10, or whether signal processing device 10 discards the higher frequency range Fu in the signal from filter bank 24 or simply leaves it unprocessed. Amplification intended to compensate for the user's hearing loss does not occur within the higher frequency range Fo.

[0056] In addition, Figure 2 The text illustrates the speech frequency range Fs, which in this example is entirely encompassed by the lower frequency range Fu. The speech frequency range Fs gives the frequency range in which the speech is located. Here, overtones are considered together. At that time, the speech frequency range Fs extends from 100Hz to 4kHz or up to 12kHz.

[0057] Here, the speech frequency range Fs is completely encompassed by the operating range Fa. However, alternative designs where the operating range Fa and the speech frequency range Fs (not shown) have the same upper and / or lower limits are also possible. Designs where the operating range Fa is greater than or less than the speech frequency range Fs are also conceivable. This is especially true in designs where the operating range Fa also includes a higher frequency range Fo, thus making the operating range Fa greater than the speech frequency range Fs.

[0058] The lower frequency range Fu has an upper limit of 26, for example, 12 kHz. Here, this upper limit 26 is, but not necessarily, the lower limit of the higher frequency range Fo.

[0059] The higher frequency range Fo has an upper limit 30, which is appropriately defined here by the technical boundary conditions of the hearing device 2, for example by the finite frequency range of the filter bank 24 and / or the finite sampling rate of the A / D converter (i.e., analog-to-digital converter) 32 that digitizes the input signal 8 from the microphone 4. In the design shown here, the higher frequency range Fo has a lower limit 28 of 12 kHz and / or an upper limit 30 of 16 kHz. This lower limit 28 is derived based on the same considerations as the upper limit 26 of the lower frequency range Fu. Here, the 16 kHz upper limit 30 is derived by the technical boundary condition of the 32 kHz sampling rate of the A / D converter 32 of the hearing device.

[0060] To further explain, Figure 3 An exemplary environment, here a cafeteria's spectrum, is shown. Frequency F is depicted vertically, and time T horizontally. The amplitude A of the corresponding frequency at a given time T is shown in grayscale. Wherein, the brighter / whiter the light, the larger the amplitude A, and conversely, the darker / blacker the light, the smaller the amplitude A. Figure 3 In the diagram, the higher frequency range Fo is enclosed in a box for easy identification, with the lower frequency range Fu immediately following below it. This box encompasses the frequency range from 11 kHz to 16 kHz. Figure 4 The image is shown in an enlarged format. Figure 3 Within the box in the image, the structure of the higher frequency range Fo is clearly visible. Numerous bright vertical stripes are clearly visible, indicating large amplitudes A appearing within a short time T, which are generated by pulsed and / or interfering noise 14.

[0061] List of reference numerals

[0062] 2 Hearing equipment

[0063] 4 microphones

[0064] 6. Sound signals

[0065] 8 Input Signals

[0066] 10. Signal Processing Device

[0067] 12 detectors

[0068] 14. Interference noise

[0069] 16 Useful noise

[0070] 18 Control Unit

[0071] 20 Output Signal

[0072] 22. Earpiece

[0073] 24 filter banks

[0074] 26 (lower frequency range) upper limit

[0075] 28 (lower limit of the higher frequency range)

[0076] 30 (upper frequency range) upper limit

[0077] 32 A / D Converter

[0078] Amplitude

[0079] F frequency

[0080] Fa's scope of work

[0081] Fg Overall Frequency Range

[0082] Fo higher frequency range

[0083] Fs Voice Frequency Range

[0084] Fu lower frequency range

[0085] M minimum amplitude

[0086] T time

Claims

1. A hearing device (2) for supplying a hearing-impaired user or a headset, - the hearing device having at least one microphone (4) which is configured to receive a sound signal (6) and to convert it into an input signal (8) in an overall frequency range (Fg), - the hearing device having a signal processing device (10) for processing the input signal (8) in a lower frequency range (Fu) which is part of the overall frequency range (Fg), - the hearing device having a detector (12) for identifying an interfering noise (14), the interfering noise (14) being a transient and / or impulse-like interfering noise (14) and having a frequency component in the lower frequency range (Fu) and in a higher frequency range (Fo) above the lower frequency range (Fu), characterized in that - the hearing device (2) is configured such that the detector (12) identifies the interfering noise (14) on the basis of the frequency component of the interfering noise (14) in the higher frequency range (Fo), - the lower frequency range (Fu) is the operating range (Fa) of the signal processing device (10) so that the signal processing device (10) limits the processing of the input signal (8) to the lower frequency range (Fu), - the hearing device (2) is configured such that the interfering noise (14) is suppressed in the lower frequency range (Fu) by means of the signal processing device (10).

2. The hearing device (2) according to claim 1, wherein the detector (12) being an impulse detector.

3. The hearing device (2) according to claim 1, wherein, the detector (12) being configured as a level detector, whereby the detector identifies the interfering noise (14) itself if the amplitude (A) of the interfering noise (14) in the higher frequency range (Fo) exceeds a minimum amplitude (M).

4. The hearing device (2) according to claim 1, wherein the detector (12) being configured as a gradient detector, whereby the detector identifies the interfering noise (14) itself if a minimum change in the amplitude (A) of the interfering noise (14) in the higher frequency range (Fo) occurs during a predefined time interval.

5. The hearing device (2) according to claim 1, wherein a speech frequency range (Fs) being completely or at least predominantly contained in the lower frequency range (Fu).

6. The hearing device (2) according to claim 1, wherein the higher frequency range (Fo) having an upper limit (30) which corresponds at least to an upper limit of a speech frequency range (Fs).

7. The hearing device (2) according to claim 1, wherein the detector being limited to the higher frequency range.

8. The hearing device (2) according to any one of claims 1 to 7, wherein, the hearing device (2) being configured such that - modifying the input signal (8) within the operating range (Fa) in order to adapt to the user.

9. Hearing device (2) according to one of claims 1 to 7, wherein the hearing device (2) is configured to - the amplification intended to compensate for the hearing impairment of the user does not take place within the higher frequency range (Fo) but only within the operating range (Fa).

10. Hearing device (2) according to one of claims 1 to 7, wherein the higher frequency range (Fo) is characterized in that there is and / or is expected to be less useful noise (16) within the higher frequency range (Fo) than within the lower frequency range (Fu).

11. Hearing device (2) according to one of claims 1 to 7, wherein the hearing device has an A / D converter (32) for digitizing the input signal (8), wherein the higher frequency range (Fo) has an upper limit (30) which corresponds at most to half the sampling rate of the A / D converter (32).

12. Method for operating a hearing device (2) for supplying a hearing-impaired user or as a headset, the hearing device having at least one microphone (4), signal processing means (10) and a detector (12), - wherein, receiving sound signals (6) within an overall frequency range (Fg) by means of the microphone (4) and converting into an input signal (8), - wherein the input signal (8) is processed by means of the signal processing means (10) within a lower frequency range (Fu), which is part of the overall frequency range (Fg), - wherein an interfering noise (14) is identified by means of the detector (12), which is a transient and / or impulse-like interfering noise (14) and which has a frequency component within the lower frequency range (Fu) and outside the lower frequency range, i.e. within a higher frequency range (Fo) above the lower frequency range (Fu), - wherein the detector (12) identifies the interfering noise (14) on the basis of the frequency component of the interfering noise (14) within the higher frequency range (Fo), - wherein the lower frequency range (Fu) is an operating range (Fa) of the signal processing means (10), so that the signal processing means (10) limit the processing of the input signal (8) to the lower frequency range (Fu), - wherein the interfering noise (14) is suppressed by means of the signal processing means (10) within the lower frequency range (Fu).

Images