Methods for directional signal processing in hearing instruments
By using a dual-input signal processing method in hearing aids, the angle and orientation of sound sources can be approximately identified, relevant sound source signals can be enhanced and irrelevant sound sources can be suppressed, solving the problem of sound signal differentiation in complex environments and improving the wearer's ability to understand conversations.
Patent Information
- Application Number
- CN202580029392.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-04-22
- Filing Date
- 2025-02-10
- Publication Date
- 2026-11-17
AI Technical Summary
Existing hearing instruments struggle to accurately distinguish between relevant and irrelevant sound signals in complex auditory environments, especially with multiple participants and high levels of noise interference, causing the voices of participants to be drowned out by background noise and other speaker signals.
The first and second input signals are generated by the two input transducers of the hearing instrument. Based on these signals, the angle and orientation of the sound source relative to the front of the wearer are approximately identified. Directional signal processing is used to enhance or reduce the sound signal components of the sound source. In particular, the relationship between the angle and orientation is used to identify the relevant sound source and enhance or suppress the signal.
It enhances the voice components of dialogue participants in complex environments in the output signal, reduces the signal components of irrelevant sound sources, and enhances the wearer's ability to understand specific dialogues.
Smart Images

Figure CN123125902A_ABST
Abstract
Description
[0001] The present invention relates to a method for directional signal processing for a hearing instrument, wherein, based on a first input signal and a second input signal of the hearing instrument, the angular direction of a sound source relative to the frontal direction of the wearer of the hearing instrument is identified at least approximately, and the input signal is processed according to the angular direction.
[0002] In hearing aids, particularly in the narrow sense, the input signal is processed into an output signal, especially through frequency band-specific signal processing, and then transmitted to the wearer's ear, for example, by means of a speaker. Here, the signal processing can be additionally tailored to the wearer and, in particular, their hearing requirements, for example, to compensate for the wearer's hearing loss in the narrow sense.
[0003] The signal processing is often performed in a directional manner, in that multiple input signals are processed into an output signal in such a way that sounds from different spatial directions of the surrounding environment are incorporated into the output signal differently, i.e., sounds from some spatial directions are suppressed, while sounds from other spatial directions are amplified.
[0004] For such directional microphones, signal processing is typically based on specific models of the environmental conditions. For example, sound sources located in front of the wearer are generally considered relevant or useful signal sources because it is assumed that the wearer will direct their gaze towards sound sources that are relevant to them. Another assumption is that sound from the wearer's rear hemisphere, depending on its spectrum, is in principle interpreted as interfering noise.
[0005] However, it is precisely in complex conversational situations with multiple participants, especially in environments with a lot of additional interfering noise (such as the so-called "cocktail party" auditory situation in a restaurant), that this signal processing reaches its limits, because the distinction between wearer-related and irrelevant sound signals is often too coarse or too imprecise. Yet, it is precisely in these complex auditory environments that such distinction is crucial for selectively highlighting specific conversational participants from background noise and / or other speakers (such as speakers participating in other conversations, i.e., those not involved with the wearer).
[0006] Therefore, the technical problem to be solved by the present invention is to provide a method that provides an additional possibility for hearing instruments, namely, to process the sound source differently according to the relevance of the sound source to the wearer of the hearing instrument.
[0007] The technical problem described herein is solved by a method for directional signal processing for a hearing instrument, wherein a first input signal is generated from ambient sound via a first input transducer of the hearing instrument, and a second input signal is generated from ambient sound via a second input transducer of the hearing instrument, wherein, based on the first and second input signals, the angular direction of the sound source relative to a first reference direction, particularly relative to the frontal direction of the wearer of the hearing instrument, is identified at least approximately, and wherein, based on the first and second input signals, the orientation direction of the sound source, particularly relative to a second reference direction, is identified at least approximately.
[0008] Here, according to the method, the acoustic signal component of the sound source is boosted or reduced in the processed signal generated based on the first and second input signals, according to the angular direction and orientation direction of the identified sound source. Advantageous and in part inventive embodiments are the technical solutions described in the dependent claims and below.
[0009] Here, hearing instruments generally include any device configured to generate an electrical input signal from ambient sound using at least one, particularly electroacoustic input transducer, process the input signal into an output signal by amplification and / or compression, particularly band-specific amplification, and generate an acoustic signal from the output signal and deliver it to the ear of the wearer, particularly using an electroacoustic output transducer (e.g., a loudspeaker, a so-called balanced armature receiver, but also a bone conduction headset). Therefore, hearing instruments particularly include headphones (e.g., "earplugs"), headphones, data glasses with loudspeakers, etc., equipped with corresponding input transducers. However, hearing instruments also include hearing aids in the narrow sense, i.e., devices for compensating for hearing loss in the wearer, in which, when processing the input signal into an output signal, the former is amplified and / or compressed, particularly band-specific amplification, in order to at least partially compensate for the wearer's hearing loss in a user-specific manner by means of an output acoustic signal generated from the output signal.
[0010] In particular, the hearing instrument can also be configured as a binaural hearing system having a first local device and a second local device, wherein in this case, the hearing instrument preferably also has (at least) two additional input transducers, wherein the first and second input transducers are arranged in the first local device, and the two additional input transducers are arranged in the second local device. The applicability of the described method is generally unrelated to this.
[0011] Here, the first or second input transducer particularly includes any of the following devices configured to generate a corresponding electrical signal from the acoustic signal. In particular, preprocessing, for example in the form of linear preamplification and / or A / D conversion, may be performed when generating the first or second input signal through the corresponding input transducer. The correspondingly generated input signal is given in particular by an electrical signal whose current and / or voltage fluctuations substantially represent the sound pressure fluctuations of the air.
[0012] Here, the angular direction of the sound source relative to the wearer's frontal orientation should be understood in particular as establishing a clear relationship between the two input transducers and the wearer's frontal orientation by wearing the hearing instrument on the head as prescribed (as is the case with a binaural hearing system), especially on or in the ear (in the case of a monoaural hearing instrument). By means of the corresponding directional processing of the two input signals, the angular direction of the sound source can thus be identified, for example, by the time difference between the corresponding signal components in the first and second input signals, wherein this angle is referenced to a first reference direction, which is given by or can be defined based on the described frontal orientation.
[0013] Here, the orientation direction of the sound source should be understood in particular as the direction along which the sound source emits its maximum sound energy, or the direction in which the maximum sound energy emitted by the sound source is located. In the case of a speaker acting as the sound source, this direction is generally equivalent to the speaker's facing direction. In the case of a loudspeaker, the orientation direction is generally given by the axis of symmetry of the diaphragm arrangement. Here, the orientation direction can be particularly referenced to a first reference direction, preferably the facing direction of the hearing instrument wearer, as a vector direction (i.e., offset towards the sound source), such that the vectors of the first and second reference directions are parallel. However, preferably, the second reference direction is given by the angular direction. By virtue of the fixed relationship between the first and second reference directions, the definition of the second reference direction can preferably be chosen according to the subsequent application, since all alternative definitions are equivalent to each other (except for the corresponding angular transformations).
[0014] Here, the approximate identification of angular direction and / or orientation direction specifically includes: pre-given multiple, at least three (and preferably more) discrete angle values as a range of possible values for the corresponding direction, determining the angle value that is closest to the actual angular direction or orientation direction, and outputting that angle value as the identified angular direction or orientation direction.
[0015] Here, at least approximate identification should be understood as the identification of the corresponding direction can be performed with the discrete angle values, or it can be performed continuously and especially precisely (within the corresponding range of possibilities, especially within the range caused by any effects of discretization, sampling rate and limited computation and storage capacity, etc.).
[0016] Here, identifying the angular direction or orientation direction based on the first and second input signals specifically includes: directly identifying on the signal components of the first and second input signals, i.e., directly applying a corresponding angle-dependent filter (e.g., a notch filter) to the first and second input signals to identify the relevant direction, and / or directly determining the running time difference of the signal components in the first and second input signals. Preferably, for such filters used for identification, the head occlusion effect can also be considered, especially by means of one or more head-related transfer functions, particularly those related to the orientation direction.
[0017] However, identifying the angular direction or orientation based on the first and second input signals also includes: applying the filter used to identify the relevant direction to the first intermediate signal and the second intermediate signal, and / or determining the running time difference of the signal components in the first and second intermediate signals. Here, the first intermediate signal is preferably derived only from the first input signal (i.e., the signal components of the second input signal do not directly enter the second intermediate signal), and the second intermediate signal is preferably derived directly from the second input signal (i.e., preferably the signal components of the first input signal do not enter the second intermediate signal).
[0018] Here, the intermediate signal can be generated by performing single-channel preprocessing on the corresponding input signal. However, the intermediate signal can also be generated by performing directional preprocessing on the first and second input signals respectively, for example, as forward and backward cardioid signals.
[0019] Here, the identification of the sound source angle direction can be performed before or together with the identification of the orientation direction.
[0020] By identifying the orientation direction of the sound source, which in particular includes the direction of maximum sound energy emission, and by considering its relationship with the angular direction, it is particularly easy to distinguish whether the sound source is a source of interference noise or a potentially useful signal source for the wearer. In particular, if only a small portion of the sound energy is emitted toward the wearer of the hearing instrument due to the orientation direction of the sound source (in the case of a given angular direction), the sound source can be interpreted as a source of interference noise. This can be distinguished, for example, by corresponding angular limits, originating from the sound source, etc. For example, the component of the sound source's orientation direction toward the wearer can in principle be boosted, or the component of the sound source's orientation direction sufficiently away from the wearer's orientation (e.g., at least perpendicular to the angular direction) can be reduced, even if the sound source (e.g., the speaker) is located in front of the wearer (i.e., in the frontal direction or within an angular range of preferably ±60°, particularly preferably ±45° around the frontal direction).
[0021] Here, preferably, the sound source is identified as a wearer-related sound source based on its orientation, and the component of the sound signal of the relevant sound source in the processed signal is enhanced. In particular, this can be done in conjunction with speech recognition (e.g., based on spectral features such as formants and / or temporal features such as characteristic pauses in the input signal), such that the sound source is identified as the speaker, and thus the relevant sound source is identified as a dialogue participant of the wearer.
[0022] Furthermore, the identification of relevant sound sources can be based on corresponding angular boundaries (starting from the sound source). For example, it can be checked whether the orientation direction relative to the angular direction (i.e., the angular direction preferably opposite to the sound source as a second reference direction) is greater than, for example, ±10°, or greater than, for example, ±22.5°, or greater than, for example, ±45° (i.e., whether the absolute value of the deviation between the orientation direction and the angular direction is greater than 10°, or 22.5°, or 45°). If this is not the case for the set boundaries, and the deviation is therefore smaller, then it is assumed that the sound source is substantially oriented towards the wearer of the hearing system and is classified as a sound source relevant to the wearer.
[0023] Advantageously, the component of the acoustic signal from the relevant sound source in the processed signal is enhanced by generating a directional signal (i.e., a first directional signal) with a corresponding central direction (i.e., a first central direction) based on the first and second input signals. This central direction is oriented away from the angular direction towards the orientation direction of the sound source, and the directional signal is incorporated into the processed signal. In particular, the processed signal is incorporated into the aforementioned output signal such that its components (which are amplified and / or compressed, if necessary, frequency-dependently) are collectively converted into output sound.
[0024] This specifically means that the first and second input signals (or the “unmixed” intermediate signals derived from them, respectively) are processed into a (first) directional signal using a directional microphone. This directional signal has a (first) central direction, which is preferably given by the direction of maximum sensitivity of its directional characteristics. Here, the directional signal is generated (i.e., for example, by using directional parameters for superimposing the input signals and / or delays for directional correlation interference) such that its central direction does not correspond to the angular direction (therefore the directional signal is not precisely oriented toward the sound source), but is slightly deflected relative to the angular direction, specifically deflected toward the orientation direction.
[0025] This is thanks to Figure 1 To illustrate, Figure 1The auditory context of the wearer 10 of the hearing instrument 1 is shown in the top view. Here, the hearing instrument 1 is provided by a binaural hearing system having a first local device 1a and a second local device 1b, each having at least one input transducer (not shown). By means of the arrangement of the two input transducers in the hearing instrument 1, a first reference direction R1 can be defined for the prescribed wearing of the hearing instrument 1, i.e., for the corresponding local devices 1a and 1b to be worn on their respective ears, which is preferably selected as the frontal direction 12 of the wearer 10.
[0026] In the angular direction α relative to the first reference direction R1, which is given here by α≈15°, and which is given by the speaker S1, the sound source 16 has an orientation direction vw1 to which its sound energy is emitted to the maximum extent. For the speaker S1, who is the sound source 16, this orientation direction vw1 is equivalent to the speaking or line of sight of the speaker S1. Here, the orientation direction vw1 is preferably defined relative to a suitably selected second reference direction R2. Here, the second reference direction R2 is selected as the (opposite) angular direction α of the speaker S1, but it can also be selected as being the same as the first reference direction R1.
[0027] By recognizing the relatively frontal position of speaker S1 (i.e., angular direction α) and its orientation direction vw1 (where vw1 ≈ 45° relative to R2) without significantly deviating from angular direction α, speaker S1 can be identified as a sound source relevant to the wearer and thus as a participant in the wearer's dialogue. Accordingly, a first directional signal D1 is generated (the directional signal, in the spatial diagram here and below, should be represented by its corresponding directional characteristics), giving it a first central direction C1 that is slightly deflected towards the orientation direction vw1 relative to the angular direction α of sound source 16 (i.e., speaker S1). Here, the first central direction C1 is the direction of maximum sensitivity of the first directional signal D1 or its directional characteristics.
[0028] In another advantageous embodiment, additionally or alternatively, a directional signal (i.e., a second directional signal) is generated based on the first and second input signals, and the component of the acoustic signal from the relevant sound source in the processed signal is enhanced by broadening the directional characteristics of the directional signal, i.e., in particular, the directional cone and other characteristics oriented relative to a reference in the angular direction, when the angular direction does not correspond to the orientation direction. The directional signal is then incorporated into the processed signal, which preferably enters the output signal.
[0029] This can specifically include: firstly processing the first and second input signals (or the "unmixed" intermediate signals derived from them, respectively) into a (second) directional signal using a directional microphone, which has a (second) central direction. Conventional signal processing would now first orient this central direction in the angular direction, i.e., directly toward the sound source. This conventional signal processing here provides a reference characteristic (having a central direction oriented in the angular direction, i.e., directly toward the sound source) and a defined broadening (which may in particular depend on the directionality and sound pressure level of the sound source signal). The current (second) directional signal now broadens the directional characteristics relative to the reference characteristics, allowing a larger portion of the sound source to enter the processed signal.
[0030] In particular, the directional characteristics of the directional signal described herein are asymmetrically broadened, with greater broadening in the direction toward the orientation direction. This is achieved by means of... Figure 2 To illustrate, Figure 2 The top view shows again the data based on... Figure 1 The same auditory context for the wearer 10 of the hearing instrument 1. In the auditory context shown, conventional signal processing, i.e., signal processing that does not consider the orientation direction wv1 of the sound source 16, will produce a reference direction signal with reference characteristics Dr (dashed line) under the same conditions.
[0031] However, considering the orientation direction wv1, it can be inferred that the main part of the speaker S1's sound is emitted to the region to the left of the angular direction α. Therefore, it is advantageous to generate a second directional signal D2 (solid line) here, whose directional characteristics have an asymmetric broadening relative to the reference characteristic Dr towards the orientation direction wv1. At the boundary of the second directional signal D2 away from the orientation direction wv1 (this boundary can be given, for example, by a defined attenuation degree relative to the maximum sensitivity, such as 3 dB or 6 dB), the second directional signal D2 can also extend along the reference characteristic Dr. In particular, through the broadening described herein, the center direction (not shown) of the second directional signal Dr can also be similar to... Figure 1 Deflected toward the orientation direction wv1.
[0032] In another advantageous embodiment, a plurality of sampling direction signals are generated based on the first and second input signals, each having a different center direction. The component of the acoustic signal from the relevant sound source in the processed signal is enhanced by weighted superposition of at least some of the sampling direction signals to generate a direction signal (i.e., a third direction signal) having its own center direction (i.e., a third center direction), such that the center direction is oriented away from the sound source's orientation direction in an angular direction. The direction signal is then fed into the processed signal, which preferably enters the output signal.
[0033] This specifically includes processing the first and second input signals (or "unmixed" intermediate signals derived from them, respectively) using a directional microphone into individual sampling direction signals, each oriented in a different direction (i.e., having a different center direction), and thus "sampling" at least one sub-region of the space (wherein the sub-region preferably contains at least the sound source). Then, by weighted superposition of at least some of these sampling direction signals, a (third) direction signal is generated, whose (third) center direction is oriented away from the angular direction toward the orientation direction of the sound source.
[0034] This is thanks to Figure 3 To illustrate, Figure 3 The top view shows again the data based on... Figure 1 The hearing instrument 1 is used in the same auditory context as the wearer 10. Here, for the angular direction α and orientation direction vw1 of the identified speaker S1, at least three sampling direction signals Da1, Da2, and Da3 (dashed lines) are generated, each with a different orientation, i.e., different center directions Ca1, Ca2, and Ca3. These center directions Ca1, Ca2, and Ca3 are given by Ca1 = 0°, Ca2 = 45°, and Ca3 = 90°. In particular, this is merely an exemplary illustration of the sampling direction signals Da1, Da2, and Da3; therefore, other sampling direction signals can also be used, and / or the center directions Ca1, Ca2, and Ca3 of the sampling direction signals Da1, Da2, and Da3 can have other angular intervals (e.g., 30° each). However, it is preferable to use at least three sampling direction signals Da1, Da2, and Da3 and / or to use sampling direction signals with ≤90° angular intervals relative to their respective adjacent center directions. The sampling direction signals Da1, Da2, and Da3 are now superimposed to generate a third direction signal D3 (solid line), which, for example, has the following form.
[0035] D3 = g1·Da1 + g2·Da2 + g3·Da3
[0036] Where g1, g2, and g3 are weighting factors and g1 + g2 + g3 = 1. The superposition is designed such that the third directional signal D3 thus has a third central direction C3, which is oriented away from the angular direction α of the sound source 16 toward its orientation direction vw1 (correspondingly, g2 > g1, g2 > g3 applies here).
[0037] Advantageously, in order to enhance the acoustic signal component of the sound source in the processed signal, frequency-dependent amplification for the upper frequency range is also additionally increased. Precisely when the orientation and angular direction of the sound source are not aligned (therefore, the sound source is not directly oriented towards the wearer of the hearing aid), under this spatial orientation, due to high directivity, only a small portion of the acoustic signal reaches the wearer of the hearing aid in the high frequency range from 1500 Hz, preferably 2000 Hz. If the sound source is a speaker (which is often the case), the loss of acoustic power in these frequency bands further increases the difficulty of speech comprehension for the wearer, because these bands contain important formants for identifying vowels and characteristic energies for consonants. Therefore, temporarily increasing the amplification in these frequency ranges may be advantageous, even if it results in a sharper or harsher overall timbre, because the focus here should outweigh the focus on the most natural possible timbre.
[0038] It has also proven advantageous to identify a sound source as irrelevant to the wearer based on its orientation, and to reduce the component of the irrelevant sound source's acoustic signal in the processed signal. In principle, the above-described embodiments for identifying relevant sound sources are similarly applicable (especially in terms of angular ranges, which are then preferably interpreted through corresponding complementary ranges, etc.) for identifying a sound source as irrelevant to the wearer. For example, a sound source located in front of the wearer (i.e., located in the frontal direction or within an angular range of preferably ±60°, particularly preferably ±45° around the frontal direction) can be interpreted as an irrelevant sound source to the wearer, and its component is correspondingly reduced when its orientation is sufficiently away from the wearer's orientation (e.g., at least perpendicular to an angular direction, which preferably indicates an orientation towards the wearer).
[0039] Advantageously, this can be achieved in particular by generating multiple sampling direction signals, each with a different center direction, based on a first input signal and a second input signal, respectively. The component of the acoustic signal from an irrelevant sound source in the processed signal is reduced by generating a direction signal with a center direction through a weighted superposition of at least some of the sampling direction signals, such that the center direction is oriented away from both the angular direction and the orientation direction of the sound source, and wherein the direction signal is incorporated into the processed signal. This implementation is particularly advantageous when the irrelevant sound source is one of several existing sound sources in the wearer's surrounding environment, and its angular direction has a relatively small angular interval (e.g., preferably ≤90°, particularly preferably ≤60°, or even ≤45°) with the angular direction of the sound source identified as relevant.
[0040] This is Figure 4 The illustration is as follows: Figure 4 The top view shows the data according to... Figure 1The hearing device 1 presents a new auditory context for the wearer 10. Here, the speaker S1, as a relevant sound source for the wearer 10, has an angular direction α of -90° relative to the wearer's facing direction 12. In this case, directional signal processing algorithms often identify the sound source as irrelevant or as interference, and weaken its signal components. However, the speaker S1 is oriented towards the wearer 10, i.e., its orientation direction wv1 coincides with the (opposite) angular direction α. In this case, the speaker S1 is identified as a relevant sound source, and its components are enhanced in the processed signal in the manner described above.
[0041] However, another speaker S2 exists in the surrounding environment of the wearer 10, with an angular direction α = -45° relative to the wearer's frontal direction 12. However, speaker S2 is not oriented towards the wearer 10, and its orientation direction wv2 is deflected by 90° relative to the first reference direction R1 given by its angular direction α. Therefore, speaker S2 participates in a dialogue with other participants (not shown) but not the wearer, and should therefore be classified as an irrelevant sound source 18.
[0042] exist Figure 4 In more complex scenarios, the two sound sources 16 and 18 are spatially close to each other (their angular distances in their respective angular directions are only 45°), yet one sound source 16 (i.e., speaker S1) is identified as relevant to the wearer 10, while the other sound source 18 (i.e., speaker S2) is identified as irrelevant. In this case, the described approach helps to reduce the component of the acoustic signal from the irrelevant sound source 18 in the processed signal by generating a third directional signal D3 with a third center direction C3 through a weighted superposition of at least some of the aforementioned sampled directional signals. This third center direction C3 is oriented away from the angular direction α and from the orientation direction wv2 of the irrelevant sound source 18, and is oriented as far as possible toward the relevant sound source 16.
[0043] In particular, the components of unrelated sound source acoustic signals in the processed signal can also be reduced by generating a directional signal with directional characteristics based on the first and second input signals. These directional characteristics have a local minimum, and preferably a global minimum, in the vicinity of the angular direction and in the direction toward the orientation direction. This specifically includes the fact that the local and preferably global minimum of the directional characteristics (which preferably also leads to the complete elimination of unrelated sound source acoustic signal components) lies within the range [0°, δ] from the angular direction α, where δ ≤ 15° and particularly preferably δ ≤ 10°, wherein the direction (i.e., whether α + δ or α - δ should be chosen) is given by the orientation direction.
[0044] Suitablely, the detection of whether the sound source is located in a first region, which is located in the front half of the wearer's space, is performed only when the sound source is located in the first region. This prevents the waste of computational power used to identify relevant sound sources when the sound source is located in a spatial region that is a priori assumed to be irrelevant to the wearer. Here, the first region can preferably be selected as a region relative to a first reference direction or the wearer's frontal direction [-67.5°, 67.5°], and particularly preferably [-60°, 60°].
[0045] Advantageously, the approximate identification of orientation direction is performed by selecting from multiple, preferably at least three, discrete core orientation directions. Here, the identification of orientation direction can also be performed in particular as follows: identifying one of three orientation regions: an orientation direction facing the wearer, an orientation direction passing in front of the wearer, and an orientation direction passing behind the wearer. For determining whether a sound source is relevant to the wearer, higher resolution is generally not required, i.e., a larger range of orientation direction values (relative to angular or reverse angular directions). Therefore, more complex determinations and, especially, more complex calculations can be avoided.
[0046] Suitably, a plurality of angle-dependent and orientation-direction-dependent filters are provided, each corresponding to an orientation direction at a specific angular direction. These filters are applied to first and second input signals and / or first and second intermediate signals, respectively derived from the first or second input signals, thereby determining at least one set of corresponding orientation-direction-dependent features, and wherein the orientation direction is identified based on these features. In particular, the filters and thus the resulting features may also depend on the corresponding angular direction.
[0047] Here, as a feature, it is preferable to determine the sound pressure level or the degree of sound attenuation, the sound originating from a sound source arranged in the relevant angular direction and oriented in the relevant orientation direction. Here, the filters are particularly constructed as notch filters, which are applied directly to the first and second input signals, or to the first and second intermediate signals derived from them respectively, and achieve spatial filtering corresponding to the relevant angular range and the corresponding orientation direction, thereby enabling the determination of the maximum boost or maximum attenuation toward one of the orientation directions.
[0048] Here, orientation-direction-related filters “scan” the space around the sound source at a given angular direction, thereby determining the orientation direction belonging to the relevant feature, especially based on the maximum or minimum value of the orientation-direction-related feature.
[0049] Here, for the degree of attenuation, it is particularly preferred to compare a comparison signal generated by applying relevant filters to the first and second input signals and / or the first and second intermediate signals with a reference signal, which preferably has omnidirectional characteristics. This comparison is particularly easy to implement, and the reference signal provides a comparison quantity for the total sound pressure level as a reference quantity for the attenuation caused by the corresponding filter.
[0050] In particular, a two-level identification process can be performed here: first, the angular direction of the sound source is determined (wherein, the determination of the angular direction can also be done in a manner different from filter-based methods, such as by means of interaural time / phase difference); then, multiple orientation-direction-related filters are provided for the identified angular direction. However, the identification of the angular direction α and the orientation direction vw can also be performed simultaneously, given multiple discrete configurations, each given by a pair (αj, vw-k), where indices j and k indicate different combinations of their respective angles. For these combinations, corresponding filters are provided, and then the angular direction and orientation direction of the sound source are identified using their respective features.
[0051] This specifically means that only a limited number of possible configurations considered relevant are examined, where each configuration is given by one of a plurality of possible angular directions and its associated orientation direction (especially relative to the corresponding angular direction or the reverse angular direction), wherein for an angular direction there may be exactly one or more orientation directions (i.e., one or more configurations under that angular direction). This allows the overhead to be kept low because only a small number of configurations are considered (significantly reduced relative to the number of potential possible configurations), and only the corresponding filters need to be provided and the associated features determined.
[0052] As another advantageous implementation, for each orientation direction under a given angular direction, an orientation-dependent head-related transfer function (ODHRTF) is provided, where each ODHRTF represents the propagation path of sound from a sound source arranged in the relevant angular direction and having the orientation direction, and where orientation-related filters are formed based on their respective ODHRTFs. This means, in particular, that the corresponding filter (e.g., as a notch filter) takes into account or preferably includes the ODHRTF from the sound source (in the case of a given angular direction and a given orientation direction) to the first or second input transducer. This ensures an accurate mapping of the spatial propagation of sound from the sound source to the corresponding input transducer, thereby minimizing spatial distortion effects in the filter that could lead to inaccurate determination of the angular or orientation direction.
[0053] Here, ODHRTF is an extension of the concept of Head Related Transfer Function (HRTF), which represents the transmission path of sound from a sound source arranged in a specific angular direction. However, unlike ODHRTF, it does not depend on the orientation of the sound source. Preferably, each ODHRTF is measured and / or simulated in calibration beforehand (i.e., prior to the method itself).
[0054] The present invention also provides a hearing instrument, comprising a first input transducer for generating a first input signal from ambient sound, a second input transducer for generating a second input signal from ambient sound, and a signal processing device, wherein the hearing instrument is configured to perform the above-described method, and wherein, in particular, the signal processing device is configured to (by correspondingly equipping a processor and a working memory addressable thereto and by means of program instructions) perform the signal processing steps of the above-described method and generate the processed signal thereon.
[0055] Preferably, the hearing instrument further includes an output transducer for generating output sound based on an output signal, wherein a signal processing device is configured to generate an output signal based on the processed signal.
[0056] The hearing device of the present invention shares the advantages of the method of the present invention. The advantages given for the method and its improvements can be similarly applied to the hearing device. Here, the hearing device can be provided in particular as a monocular, i.e., a one-piece device, which is preferably worn by the wearer on one ear. In particular, the hearing device can also be provided as a binaural hearing system with two local devices, wherein each individual device is worn on one or the other ear of the wearer.
[0057] The embodiments of the present invention will now be described in more detail with the aid of the accompanying drawings. The embodiments are illustrated herein as follows: Figure 1 The diagram illustrates the auditory context of a hearing aid wearer, featuring a nearly frontal conversation participant and corresponding directional signals from the hearing aid depending on the participant's orientation. Figure 2 Showing according to Figure 1 The hearing context of the wearer and the broadening of directional signals allow for better acquisition of the speech components of the dialogue participants. Figure 3 Showing according to Figure 1 The wearer in the auditory context and the generation of directional signals based on multiple pre-given sampling directional signals to acquire the speech components of the dialogue participants. Figure 4 Showing according to Figure 1 The wearer of the hearing device in auditory situations involving lateral conversational participants and speakers unrelated to them, and Figure 5 The block diagram shows the basis Figure 1 One possible implementation of a hearing aid.
[0058] In all the accompanying drawings, the corresponding parts and dimensions are given the same reference numerals.
[0059] exist Figure 5 In the diagram, a hearing instrument 1 is schematically shown, which is given herein as a binaural hearing system 3 having a first local device 1a and a second local device 1b. Therefore, the binaural hearing system 3 corresponds to... Figures 1 to 4 The hearing instrument. Here, the first local device 1a has a first input transducer M1, and the first and second local devices 1b have a second input transducer M2. Here, the first input transducer M1 and the second input transducer M2 are respectively provided by corresponding microphones. In particular, the first and / or second local devices 1a, 1b may also have one or more other microphones or other input transducers.
[0060] A first input transducer M1 is configured to generate a first input signal E1 from ambient sound 2 during operation of the hearing instrument 1. Correspondingly, a second input transducer M2 is configured to generate a second input signal E2 from ambient sound 2 during operation of the hearing instrument 1. The first input signal E1 and the second input signal E2 are fed to a signal processing device 4 of the binaural hearing system 3, where the two input signals E1 and E2 are processed into a first or a second output signal A1 and A2.
[0061] Here, the signal processing apparatus 4 specifically includes a first signal processing unit 4a, which is arranged in the first local device 1a and directly receives the first input signal E1, and a second signal processing unit 4b, which is arranged in the second local device 1b and directly receives the second input signal E2. Furthermore, the second input signal E2 is transmitted to the first local device 1a (with data compression if necessary) via communication units (not shown, which are respectively arranged in the first and second local devices 1a and 1b) configured and designed accordingly, so that the second input signal E2 can be processed together with the first input signal E1 in the first signal processing unit 4a. Similarly, the first input signal E1 (with data compression if necessary) can be transmitted to the second local device 1b for processing in the second signal processing unit 4b.
[0062] When processing the first and second input signals E1, E2 into the first or second output signals A1, A2, frequency band-specific amplification and / or compression are performed, wherein the signal processing from the two input signals E1, E2 to the first or second output signals A1, A2 is particularly performed in a direction-dependent manner, that is, components of individual sound sources from different spatial directions in the ambient sound can be amplified with different intensities. Furthermore, the signal processing is particularly tailored to the hearing requirements of the wearer of the hearing instrument 1.
[0063] The first local device 1a also has a first output transducer L1, which is configured to generate a first output sound 6a from a first output signal A1. Similarly, the second local device 1b has a second output transducer L2, which is configured to generate a second output sound 6b from a second output signal A2.
[0064] Figure 5 The schematic diagram of the hearing instrument 1 here shows the corresponding local devices 1a and 1b as so-called behind-the-ear (BTE) hearing aids, which have an associated earpiece in which a corresponding output transducer L1 or L2 is arranged. However, the local devices 1a and 1b can also be designed in other structural forms, in particular as in-the-ear (ITE), in-the-canal (ITC), completely in-the-canal (CIC), receiver-in-the-canal (RIC) hearing aids, or especially as ear-worn headphones that are not exclusively or primarily used to compensate for hearing loss. In particular, the hearing instrument 1 can also be designed as a monophasic hearing aid.
[0065] Although the present invention has been described and illustrated in more detail through preferred embodiments, the invention is not limited to the disclosed examples, and those skilled in the art can derive other variations therefrom without departing from the scope of protection of the invention.
[0066] List of reference numerals
[0067] 1. Hearing instruments
[0068] 1a / b First / Second Local Devices for Hearing Instruments
[0069] 2. Ambient sound
[0070] 3. Binaural hearing system
[0071] 4. Signal processing device
[0072] 4a / b First / Second Signal Processing Unit
[0073] 6a / b First / Second Output Sound
[0074] 10 Wearers
[0075] 12. Front view
[0076] 16 sound sources
[0077] A1 / 2 First / Second Output Signals
[0078] C1 / 2 / 3 First / Second / Third Center Direction
[0079] Ca1-3 Sampling Direction: Center Direction of Signal
[0080] D1 / 2 / 3 First / Second / Third Directional Signals
[0081] Da1-3 Sampling Direction Signal
[0082] Dr Reference Characteristics
[0083] E1 / E2 First / Second Input Signals
[0084] L1 / 2 First / Second Output Transducers
[0085] M1 / 2 First / Second Input Transducer
[0086] R1 / 2 First / Second Reference Direction
[0087] S1, S2 Speakers
[0088] vw1, vw2 orientation direction
[0089] α Angle direction
Claims
1. A method for directional signal processing in a hearing instrument (1), - wherein, A first input signal (E1) is generated from ambient sound (2) by the first input transducer (M1) of the hearing instrument (1), and a second input signal (E2) is generated from ambient sound (2) by the second input transducer (M2) of the hearing instrument (1). - Wherein, the angular direction (α) of the sound source (16, 17) relative to the first reference direction (R1), especially relative to the frontal direction (12) of the wearer (10) of the hearing instrument (1), is identified at least approximately based on the first input signal (E1) and the second input signal (E2). - Wherein, the sound sources (16, 18) are identified at least approximately based on the first input signal (E1) and the second input signal (E2), particularly their orientation directions (vw1, vw2) relative to the second reference direction (R2), and - Wherein, based on the identified angular direction (α) and the identified orientation direction (vw1, vw2) of the sound source (16, 12), the components of the sound signal of the sound source (16, 18) are boosted or reduced in the processed signal generated based on the first input signal (E1) and the second input signal (E2).
2. The method according to claim 1, wherein The sound source (16, 18) is identified as a sound source (16) relevant to the wearer (10) based on the orientation direction (vw1, vw2) of the sound source (16, 18), and the component of the sound signal of the relevant sound source (16) is enhanced in the processed signal.
3. The method according to claim 2, wherein In the signal processing, the component of the acoustic signal of the associated sound source (16) is boosted by generating a directional signal (D1-3) with a corresponding center direction (C1-3) based on the first input signal (E1) and the second input signal (E2), the center direction being oriented away from the angular direction (α) toward the orientation direction (vw1) of the associated sound source (16), and The direction signals (D1-3) are incorporated into the processing signals.
4. The method according to claim 2 or claim 3, wherein Direction signals (D1-3) are generated based on the first input signal (E1) and the second input signal (E2), and In the signal processing, the component of the acoustic signal of the relevant sound source (16) is enhanced by broadening the directional characteristics of the directional signal (D2) relative to the reference characteristic (Dr) oriented in the angular direction (α) when the angular direction (α) does not correspond to the orientation direction (vw1). The direction signal (D2) is included in the processing signal.
5. The method according to claim 4, wherein The directional characteristics of the directional signal (D1-3) are asymmetrically broadened, and are broadened even more in the direction of the orientation direction (vw1).
6. The method according to any one of claims 2 to 5, wherein Multiple sampling direction signals (Da1-3) are generated based on the first input signal (E1) and the second input signal (E2), respectively. These multiple sampling direction signals each have a different center direction (Ca1-3). In the signal processing, the component of the acoustic signal of the relevant sound source (16) is boosted by generating a directional signal (D3) with a corresponding center direction (C3) through weighted superposition of at least some of the sampled directional signals (Da1-3), such that the center direction (C3) is deflected from the angular direction (α) toward the orientation direction (vw1, vw2) of the relevant sound source (16), and The direction signal is incorporated into the processing signal.
7. The method according to any one of claims 2 to 6, wherein, Additional frequency-dependent amplification is provided for the upper frequency range.
8. The method according to claim 1, wherein The sound source (16, 18) is identified as an irrelevant sound source (18) to the wearer (10) based on the orientation direction (vw1, vw2) of the sound source (16, 18), and the component of the sound signal of the irrelevant sound source (18) is reduced in the processed signal.
9. The method according to claim 8, wherein, A first number of sampling direction signals (Da1-3) are generated based on the first input signal (E1) and the second input signal (E2), respectively. These signals have different center directions (Ca1-3), and... The method for reducing the acoustic signal components of the unrelated sound source (18) in the processed signal is to generate a directional signal (D3) with a central direction (C3) by weighted superposition of at least some of the sampled directional signals (Da1-3), such that the central direction (C3) is oriented away from the angular direction (α) and from the orientation direction (vw2) of the unrelated sound source (18), and The direction signal is incorporated into the processing signal.
10. The method according to any one of the preceding claims, wherein Multiple angle-dependent and orientation-direction-dependent filters are provided, each corresponding to an orientation direction (vw1, vw2) at a specific angle direction (α). These filters are applied to first and second input signals (E1, E2) and / or to a first and a second intermediate signal derived from the first or second input signal (E1, E2), thereby determining at least one set of corresponding orientation-direction-dependent features. Among them, the orientation direction (vv1, vw2) is identified based on the orientation direction related features.
11. The method according to claim 10, wherein, Provide a head orientation-dependent transfer function for each orientation direction (vv1, vw2) under a given angular direction (α). Wherein, each head-direction-related transfer function represents the sound transmission path from the sound source (16, 18), which is arranged in the relevant angular direction (α) with the orientation directions (vv1, vw2), and The filters are formed based on their respective head-direction-related transfer functions.
12. A hearing instrument (1), comprising - A first input transducer (M1) is used to generate a first input signal (E1) from ambient sound (2). - A second input transducer (M2) for generating a second input signal (E2) from ambient sound (2), and - Signal processing device (4). wherein The hearing instrument (1) is configured to perform the method according to any one of the preceding claims.
