close

Вход

Забыли?

вход по аккаунту

?

DESCRIPTION JP2015039208

код для вставкиСкачать
Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JP2015039208
A binaural hearing aid system is provided that compensates for the loss of speech
comprehension under noise. SOLUTION: Microphones 14, 16, 14B, 16B providing microphone
sound signals 18, 20, 18B, 20B in response to voice, and estimation values of target signal 26
and noise signal 30 based on microphone sound signals Signal separation apparatus 12, a phase
shift circuit for shifting the phase of one estimated value of the target signal and the noise signal,
and one of the original target signal and the noise signal by phase shift of the estimated value of
the one of the target signal and the noise signal. A phase-shift adder 50, the first receiver 48
converting the receiver input signal into an acoustic signal transmitted towards the eardrum of
one of the users of the binaural hearing aid system 10, the receiver input signal And a second
receiver 48B that converts the signal into an acoustic signal that is transmitted towards the
user's other tympanic membrane. [Selected figure] Figure 7
Hearing aid with signal enhancement function
[0001]
A novel binaural hearing aid system is provided that compensates for the loss of speech
comprehension ability under the noise of a hearing impaired user.
[0002]
Deaf people often face at least two different problems.
11-04-2019
1
Hearing loss, which is an increase in the hearing threshold level, and loss of understanding of
advanced speech in noise compared to normal hearing people. For the majority of hearing
impaired patients, the results of the speech comprehension test under noise are inferior to those
with normal hearing, even if the audibility of the incoming sound is restored by amplification.
The individual's speech listening threshold (SRT) is the signal to noise ratio needed in the signal
presented to achieve a word recognition correct answer rate of 50 percent in a hearing test
under noise.
[0003]
Current digital hearing aids that use multi-channel amplification and compression signal
processing can provide amplified sound to the hearing impaired and can easily restore their
audibility. The patient's listening ability is thus improved by making the conversation cues
previously unheard audible.
[0004]
Loss of speech comprehension under noise is currently the most important task for most hearing
aid users. The conventional way of increasing the SRT of a hearing aid is to apply either
beamforming or spectral subtraction techniques.
[0005]
In the former, at least one microphone in combination with a number of fixed or adaptive filters
is used to enhance the signal from the estimated object direction and simultaneously suppress all
other signals.
[0006]
The purpose of the spectral subtraction technique is to estimate the long-term noise spectrum to
reduce the gain of the frequency band where the instantaneous target signal strength is lower
than the long-term noise strength.
Although these methods are quite different from the technical point of view, they have a common
11-04-2019
2
purpose of enhancing the signal of interest and removing noise disturbances.
[0007]
These methods can not take into account the listener's intention, and may remove part of the
audio signal that the listener is trying to concentrate.
[0008]
The following discloses a novel method of emphasizing the desired signal.
This novel method exploits the human auditory system's ability to focus on the desired signal.
Also disclosed is a novel binaural hearing aid system that utilizes this novel method.
[0009]
Listening in complex sound fields is facilitated to a considerable extent by the binaural
processing of the auditory system. Cues occur in the sound field due to diffraction effects by the
auricle, concha, head and body, and reflection effects in a reverberant environment. These are
very specific to each subject.
[0010]
The most important clues in binaural processing are interaural time difference (ITD) and
interaural level difference (ILD). The ITD results from the difference in distance from the source
to both ears. This clue is basically useful up to about 1.5 kHz, beyond which the auditory system
can not resolve the ITD clue.
[0011]
The above interaural level difference is the result of diffraction and is determined by the relative
position of the ear to the sound source. Although this clue predominates above 2 kHz, the
11-04-2019
3
auditory system is equally sensitive to changes in ILD throughout the spectrum.
[0012]
It has been argued that deaf people benefit most from ITD tactics, as lower frequencies tend to
diminish the degree of hearing loss.
[0013]
Improving conversational comprehension by manipulating the interaural relative phase and
interaural relative level of the target signal, ie the signal the listener wants to hear and the noise
signal, ie the signal the listener feels disturbed It is shown to do.
This is as if the auditory system is adapted to separate out the different ITD and ILD coded
signals and perform natural noise suppression, in order to make it easier to focus on the signal of
interest. is there.
[0014]
In both ears, when the signal of interest is presented in antiphase, ie 180 ° out of phase, and the
noise signal is in phase, compared to when both signals are presented in phase to both ears, It
has been found that the binaural masking level difference (BMLD) can be increased by 13 dB. It
is also possible to improve the BMLD by 20 dB depending on the type of noise.
[0015]
In the opposite situation where the phase of the noise signal is shifted and the signal of interest
is presented in phase, there is a slight degradation in performance.
[0016]
In the novel method, the user of the binaural hearing aid system such that at least one of the
signal of interest and the noise signal is estimated and the at least one estimate improves the
user's speech comprehension under noise To be presented.
11-04-2019
4
[0017]
For example, the listener will hear a voice with a signal S that the listener wants to hear and a
noise N that the listener feels disturbing.
That is, the audio signal is S + N.
The desired signal S is estimated based on the audio signal S + N. The estimated value is denoted
as ES. By subtracting the estimate ES twice from the speech signal S + N, the result is a modified
signal S + N-ES-ES. Since ES is approximately equal to S, the result of the subtraction is
approximately N-ES, which is approximately equal to -S + N, ie the original audio signal with the
desired signal S replaced by a 180 ° phase shifted signal S. And, in order to improve BMLD and
SRT, the original signal S + N is presented to one ear of the user and the phase shifted signal NES, more precisely S + N-2 ES, is presented to the other ear.
[0018]
Alternatively, to improve BMLD and SRT, both the desired signal S and the noise signal N are
estimated, and the sum ES + EN of the estimates is presented to one ear of the user and phaseshifted sum-- ES + EN may be presented to the other ear.
[0019]
In order to improve BMLD and SRT, the desired signal S and the noise signal may be
interchanged to shift the phase of the noise signal estimate instead of the desired signal, but in
comparison with the case where the desired signal S is shifted in phase. Performance is reduced.
[0020]
The noise is ambient conversational sound, the sound of tableware at a restaurant, music (if the
conversational sound is a desired signal), traffic noise, etc.
[0021]
The purpose of the above method is to provide the user's auditory system with natural noise
suppression and to isolate the signal of interest from the noise signal by presenting the signals
rather than removing some of the signals. It is to be.
11-04-2019
5
[0022]
In this case, if the target signal and the noise signal are switched due to any reason (for example,
the direction of the estimated target is incorrect or the device can not perform sufficient target /
noise separation), the target signal Although its performance can be emphasized, its performance
is slightly reduced.
[0023]
This is impossible with conventional noise suppression techniques.
The reason is that the target signal is assumed to be noise in this case and is suppressed.
[0024]
And, a novel binaural hearing aid system is provided.
The binaural hearing aid system comprises an estimate of one of a target signal and a noise
signal based on at least one microphone providing at least one microphone audio signal in
response to received speech, and the at least one microphone audio signal. A signal separation
device configured to provide a phase shift circuit configured to shift the phase of an estimate of
one of the target signal and the noise signal, and a phase representing speech received by the at
least one microphone A phase shift adder connected to provide a shifted signal, wherein an
estimate of one of the target signal and the noise signal substantially replaces one of the original
target signal and the noise signal. The phase shift adder and the receiver input signal are
transmitted to the tympanic membrane of one of the users of the binaural hearing aid system.
And it includes a first receiver for converting the acoustic signal, a second receiver for converting
into acoustic signals transmitted toward the receiver input signal of the user to the other of the
eardrum.
The receiver input of one of the first and second receivers is connected to a signal representing a
phase shifted signal.
11-04-2019
6
The other receiver input of the first receiver and the second receiver is connected to a signal
representative of the sound received by the at least one microphone.
[0025]
In addition, a novel method is provided to enhance binaural signals in a binaural hearing aid
system.
The method comprises the steps of providing at least one microphone audio signal in response to
speech; providing an estimate of one of a target signal and a noise signal based on the at least
one microphone audio signal; Shifting the phase of the estimate of one of the signal and noise
signals, and providing a phase shifted signal representing the at least one microphone audio
signal, wherein the phase of one of the target signal and the noise signal is The step of replacing
one of the original target signal and the noise signal by the shifted estimate, and transmitting a
signal representing the phase shifted signal to the eardrum of one of the users of the binaural
hearing aid system Sending a signal representative of the at least one microphone audio signal to
the other eardrum of the user And it includes the step of.
[0026]
If the estimate of one of the object signal and the noise signal is equal to the original one of the
corresponding object signal and the noise signal, the phase shifted estimate can be exactly
replaced by the original signal. However, usually, the estimate of the signal deviates from the
original signal, and replacement of the original signal with the estimate does not replace that
deviation. Therefore, it is expressed that the original signal is substantially replaced by the
estimated value.
[0027]
Throughout the specification, it can be said that one signal represents another signal if, for
example, one signal is a function of the other signal, such as analog to digital conversion of the
other signal or If it can be formed by digital / analog conversion, or if one signal can be formed
by conversion from another acoustic signal to an electronic signal or vice versa, or one signal is
formed by analog or digital filtering or mixing of the other signal If possible, or if one signal can
be formed by conversion such as frequency conversion of another signal.
11-04-2019
7
[0028]
Furthermore, the particular circuit, eg the signal processed by the signal processor, is any analog
or digital signal that forms part of the signal path from the source of the signal of interest to the
input to the circuit of interest, such as the signal processor. It may be identified by the name used
for identification of
For example, the output signal of the microphone, ie the name microphone sound signal, is any
analog or digital signal, including a preprocessed microphone sound signal which forms part of
the signal path from the output of the microphone to the input of the signal processing device It
may also be used to identify
[0029]
At least one microphone may comprise a single microphone, but preferably, at least one
microphone comprises two microphones. Furthermore, at least one microphone may have more
than two microphones to improve the separation of the object signal and the noise signal.
[0030]
To improve signal enhancement, the second hearing aid may also include at least one
microphone that provides a microphone audio signal in response to the received audio. In this
case, the transceiver of the first hearing aid is connected to receive a signal representative of the
microphone audio signal of the second hearing aid, and the signal separation device is targeted
based on the audio signals of the first and second hearing aids. It is configured to provide an
estimate of the signal and an estimate of the noise signal.
[0031]
Preferably, the phase shift circuit shifts the phase of the estimate of the object signal, preferably
the phase shift is in the range of 150 ° to 210 °, more preferably the phase shift is equal to
about 180 °. Most preferably equal to 180 °.
11-04-2019
8
[0032]
The signal separation device may be configured to provide an estimate based on the spectral
characteristics of the audio signal, as is known in the art of noise suppression.
However, according to the novel method, the noise signal estimate is not suppressed at the
output presented to the user, but rather the target signal estimate and the noise signal estimate
improve the user's SRT. Presented to the user at
[0033]
The signal separation device may be configured to provide an estimate based on statistical
characteristics of the speech signal, as known in the art of noise suppression. However, according
to the novel method, the noise signal estimate is not suppressed at the output presented to the
user, but rather the target signal estimate and the noise signal estimate are used in a manner to
improve the user's SRT. Presented to
[0034]
The signal separation device may include a beamformer, which may be configured to provide an
estimate based on the microphone sound signals of the first and second hearing aids. The
beamformer of the signal separation device is not suppressed at the output at which the noise
signal estimate is presented to the user, but rather the target signal estimate and the noise signal
estimate improve the user's SRT It differs from conventional beamformers in that it is presented
to the user at
[0035]
The beamformer combines the microphone audio signals output by the plurality of microphones
of at least one microphone into a target signal whose sensitivity to the sound source in different
directions changes with respect to the plurality of microphones. Throughout the specification,
directional patterns are represented by plots of sensitivity change as a function of direction.
Typically, the directivity pattern has at least one direction in which the microphone signals
11-04-2019
9
substantially cancel one another. Such directions are referred to as null directions throughout the
specification. The directivity pattern may include several null directions depending on the
number of microphones in the plurality of microphones and on signal processing.
[0036]
The beamformer may be a fixed beamformer having a directional pattern fixed with respect to
the head of the user. The beamformer may, for example, be based on at least two microphones,
the directivity pattern having a maximum in the forward direction of the user, ie in the direction
in which the user looks forward, and in the opposite direction, ie backwards of the user Has a
null direction.
[0037]
The beamformer may be based on more than two microphones and may include microphones of
a binaural hearing aid using wireless or wired communication technology. The wideness of the
distance between the microphones can be used to form a directional pattern with narrow beams
to improve the spatial separation of the target signal estimate from the noise signal estimate. The
noise signal estimate can be obtained by subtracting the target signal estimate from the
microphone speech signal of one of the plurality of microphones using the conventional output
of the beamformer as the target signal estimate.
[0038]
If the microphones of both hearing aids of the binaural hearing aid system cooperate with the
beamformer, each microphone signal must be sampled substantially synchronously. A small
deviation of as much as 20-30 μS of the sampling time of each microphone signal of the two
hearing aids may result in a non-negligible deviation in beam direction. Furthermore, when the
hearing aid is operated asynchronously, the time lag of the sampling time of each microphone
signal necessarily changes slowly with time, resulting in an acoustic beam whose direction is
alternated. .
[0039]
Thus, the hearing aids of the binaural hearing aid system may be synchronized, for example as
11-04-2019
10
disclosed in more detail in WO 02/07479.
[0040]
The beamformer may include an adaptive filter configured to filter each microphone audio signal
and to adapt each filter coefficient for adaptive beamforming directed to a source.
For example, the beamformer may be adapted to optimize the signal to noise ratio.
[0041]
The adaptive beamformer allows to focus on moving sound sources or to focus on stationary
sound sources when the user of the hearing aid system is moving. Furthermore, the adaptive
beamformer is adaptable to changes in the audio environment, such as the appearance of new
sound sources, the disappearance of noise sources, or the movement of noise sources to the user
of the hearing aid system.
[0042]
An adaptive beamformer can be designed under the assumption that the signal received at the at
least one microphone can be modeled as a combination of noise plus the signal of interest from a
given direction of interest.
[0043]
[0044]
Here, h i (n) is an impulse response of sound propagation from the sound source emitting the
signal s (n) to the ith microphone, and v i (n) is a noise signal in the same microphone.
The noise signal may include both directional noise and other types of noise, such as diffuse
noise or bubble noise.
11-04-2019
11
[0045]
The filter coefficients can be determined adaptively by solving the following optimization
problem.
[0046]
[0047]
The solution of the optimization problem can be derived adaptively using the least mean square
method, the recursive least square method, the steepest descent method, or some other type of
numerical optimization algorithm.
[0048]
Once the estimate of the signal of interest and the estimate of the noise signal are determined,
the signal is presented to the user in a manner that improves the user's SRT.
[0049]
Preferably, in the user's ears, the estimates of the signal of interest are presented in antiphase, ie
180 ° out of phase with each other, and the estimates of the noise signal are presented in phase.
Then, in the first hearing aid, the first adder is connected to the signal separation device, and
outputs the sum of the estimation value of the target signal provided by the signal separation
device and the estimation value of the noise signal, The output of the adder is connected to a
signal processor for further processing, eg hearing loss compensation, and the output of the
signal processor is connected to an output transducer which outputs an output corresponding to
one ear of the user The output of the first or first adder is directly connected to the output
transducer.
A second adder is connected to the signal separator and outputs the sum of the inverse of the
estimate of the target signal and the estimate of the noise signal provided by the signal separator,
the output of the second adder being Connected to the transceiver, the transceiver transmits the
11-04-2019
12
output of the second summer to the other hearing aid, the other hearing aid comprises a
transceiver for receiving the output of the second summer.
The output of the transceiver is connected to a signal processor for further processing, eg
hearing loss compensation, and the output of the signal processor is connected to an output
transducer which outputs an output corresponding to the other ear of the user Or the output of
the transceiver is directly connected to the output transducer.
[0050]
Alternatively, although the improvement in the user's SRT is reduced to some extent, in the user's
ears, the estimated values of the noise signal are presented in antiphase, that is, 180 ° out of
phase with each other, and the estimated value of the target signal is It may be presented in the
same phase.
[0051]
Preferably, the first hearing aid comprises a delay between the adder and the main transducer,
and the relative phase of the signals output by the respective output transducers of the first
hearing aid and the second hearing aid is maintained Ru.
[0052]
While the improvement in SRT as a function of phase shift is greatest at 180 °, the function has
a flat maximum value such that the improvement obtained by phase shift in the range of 150 °
to 210 ° is close to the greatest improvement. It has a sine wave shape.
Thus, the phase shift does not necessarily have to be exactly 180 °, but has values in the range
of 135 ° to 225 °, more preferably 150 ° to 210 °.
[0053]
The novel binaural hearing aid system may include a multi-channel first hearing aid in which the
microphone audio signal is divided into multiple frequency channels.
[0054]
11-04-2019
13
Correspondingly, individual target signal estimates and noise signal estimates may be provided
for each frequency channel of the plurality of frequency channels, or provided for one or more
selected frequency channels of the plurality of frequency channels. Or one or more estimates of
the signal of interest and noise signal estimates may be provided for each group of one or more
of the selected frequency channels of the plurality of frequency channels, or Signal estimates and
noise signal estimates may be provided based on all frequency channels of the plurality of
frequency channels.
[0055]
The plurality of frequency channels may include warp frequency channels, eg, all of the
frequency channels may be warp frequency channels.
[0056]
The novel binaural hearing aid system further comprises circuitry used in other prior art hearing
loss compensation methods so that the novel circuitry or other prior art circuitry works properly
in various types of audio environments. It may be configured to be selectable.
Various voice environments include speech sounds, loud speech sounds, dishes in the restaurant,
music, traffic noises, etc.
[0057]
The novel binaural hearing aid system may include, for example, a digital signal processor (DSP),
the processing of which is controlled by selectable signal processing algorithms, each algorithm
adjusting the actual signal processing to be performed It has parameters.
An example of such a parameter is the gain of each frequency channel of a multi-channel hearing
aid.
[0058]
One of the selectable signal processing algorithms operates according to a novel method.
11-04-2019
14
[0059]
For example, various algorithms may provide prior art noise suppression, ie, attenuation of
unwanted signals and amplification of the desired signal.
[0060]
Microphone audio signals obtained from different audio environments may have very different
features, such as average and maximum sound pressure levels (SPL) and / or frequency
components.
Thus, a specific program is associated with each type of speech environment, and the setting of
the algorithm parameters of the particular signal processing algorithm can provide the bestprocessed speech of the particular speech environment .
Such a set of parameters typically controls parameters related to wideband gain, corner
frequency or slope of the frequency selective filter algorithm, eg knee point and compression
ratio of automatic gain control (AGC) algorithm Parameters may be included.
[0061]
The signal processing characteristics of each algorithm can be determined when performing an
initial fitting operation at the hearing aid dealer and can be programmed into the non-volatile
memory area of the novel binaural hearing aid system.
[0062]
The novel binaural hearing aid system may have a user interface, such as a button on the hearing
aid housing, a toggle switch, etc, or a remote control, whereby the user of the novel binaural
hearing aid system has his own audio environment The user can select one of the available signal
processing algorithms to obtain the desired hearing loss compensation.
[0063]
The new binaural hearing aid system automatically classifies the user's voice environment into
one of many voice environment categories such as conversational sounds, loud conversational
sounds, tableware sounds at restaurants, music, traffic noise, etc. Then, appropriate signal
11-04-2019
15
processing algorithms well known in the art may be selected automatically.
[0064]
Hereinafter, preferred embodiments of the present invention will be described in more detail
with reference to the drawings.
[0065]
1 schematically shows an example of a novel binaural hearing aid system.
1 schematically shows an example of a novel binaural hearing aid system.
1 schematically shows an example of a novel binaural hearing aid system.
1 schematically shows an example of a novel binaural hearing aid system.
1 schematically shows a signal separation device with an adaptive beamformer based on two
microphones.
1 schematically shows a signal separation device based on four microphones.
1 schematically shows an example of a novel binaural hearing aid system.
[0066]
The present invention will be described more fully hereinafter with reference to the
accompanying drawings, which illustrate exemplary embodiments of the present invention.
The invention can, however, be embodied in different forms and should not be considered as
11-04-2019
16
limited to the embodiments disclosed herein.
Rather, these embodiments are provided so that this disclosure will be exhaustive and complete,
and will fully convey the scope of the invention to those skilled in the art. Like reference
numerals refer to like elements throughout the drawings. Therefore, the same elements will not
be described in detail in the description of each figure.
[0067]
An example of a novel binaural hearing aid system 10 is schematically shown in FIG.
[0068]
The novel binaural hearing aid system 10 comprises first and second hearing aids 10A, 10B.
The second hearing aid 10B includes a receiver 48B and a transceiver (not shown) that receives
an input signal from the first hearing aid 10A to the receiver 48B by wired or wireless
transmission. Thus, in the example shown, the sound output signal emitted from the second
hearing aid 10B is controlled by the first hearing aid 10A.
[0069]
The first hearing aid 10A includes one microphone 14 and provides a microphone sound signal
18 in response to the sound received by the microphone 14. The microphone audio signal 18
may be pre-filtered at each pre-filter (not shown) known in the art and then input to the signal
separation device 12. The signal separation device 12 estimates a target signal, subtracts the
estimated target signal from the microphone sound signal 18 twice, and acquires a signal which
will be hereinafter referred to as a “phase shifted signal”. The signal represents the
microphone audio signal 18, but the original target signal has been replaced by an estimate of
the target signal 180 ° phase shifted. The phase shifted signal is output to the transceiver (not
shown) of the first hearing aid 10A for transmission to the second hearing aid 10B. The receiver
48 of the first hearing aid 10A converts the microphone audio signal 18 into an acoustic signal
that is transmitted to the tympanic membrane of one of the user's ears. The receiver 48B of the
second hearing aid 10B converts the phase shifted signal into an acoustic signal that is
transmitted to the tympanic membrane of the user's other ear. This improves BMLD and SRT. The
11-04-2019
17
signal separation device 12 may be configured to provide estimates based on statistical
properties of the time domain, spectrum, and / or microphone speech signal, as is known in the
art of noise suppression. In addition, prior to the input to the respective receiver 48, 48B, further
processing may be performed on each signal, for example to provide hearing loss compensation
for each signal.
[0070]
The novel binaural hearing aid system 10 shown in FIG. 2 is based on the microphone audio
signal 18 which may be pre-filtered by the signal separation device 12 shown in FIG. It is similar
to the hearing aid system shown in FIG. 1 except that it is configured to provide both of the
estimates of.
[0071]
The estimated value of the target signal 26 is added to the estimated value of the noise signal 30
in the first adder 42 and the output sum of the estimated value of the target signal 26 and the
estimated value of the noise signal 30 is input to the output transducer 48 Be done.
The output transducer 48 converts the output of the first summer 42 into an acoustic output
signal that is transmitted to the tympanic membrane of the user wearing the binaural hearing aid
system 10.
[0072]
Furthermore, the estimate of the signal of interest 26 is subtracted from the estimate of the noise
signal 30 in the second adder 50, corresponding to a phase shift of 180 °. The output of the
second summer 50 is sent to the output transducer 48B and converted to an acoustic output
signal that is transmitted to the other eardrum of the user wearing the binaural hearing aid
system 10. This improves BMLD and SRT.
[0073]
The estimate of the target signal 26 and the estimate of the noise signal 30 can be interchanged,
11-04-2019
18
so instead of shifting the phase of the estimate of the target signal 26, instead of presenting the
noise signal 30 before presenting it to one of the user's eardrum. The phase of the estimate can
be shifted by 180 °. The resulting improvements in BMLD and SRT are less than the gains
achieved by phase shifting the estimate of the target signal 26.
[0074]
In the novel binaural hearing aid system 10 shown in FIG. 3, the microphone audio signal 18B
output from the microphone 14B of the second hearing aid 10B is transmitted to the first
hearing aid 10A by wired or wireless transmission, and the signal separation device 12 To the
hearing aid system shown in FIG. Thus, for example, with beamforming as described in more
detail below, the signal separation device 12 can provide an estimate of the target signal 26
based on both of the microphone audio signals 18, 18B. When the user wears the first and
second hearing aids 10A and 10B at predetermined positions in their own ears, the distance
between the microphones 14 and 14B is relatively large, so a narrow beam is formed and the
target signal is It is possible to achieve good spatial separation from the noise signal.
[0075]
The novel binaural hearing aid system 10 shown in FIG. 4 is a microphone audio signal 18 in
which the signal separation device 12 shown in FIG. 4 may be pre-filtered in the same manner as
the signal separation device shown in FIG. Based on that, it is similar to the hearing aid system
shown in FIG. 3 except that it is configured to provide both an estimate of the object signal 26
and an estimate of the noise signal 30.
[0076]
The estimated value of the target signal 26 and the estimated value of the noise signal 30 are
added in the first adder 42, and the output sum of the estimated value of the target signal 26 and
the estimated value of the noise signal 30 is input to the output transducer 48 Be done.
The output transducer 48 converts the output of the first summer 42 into an acoustic output
signal that is transmitted to the tympanic membrane of the user wearing the binaural hearing aid
system 10.
11-04-2019
19
[0077]
Furthermore, the estimate of the signal of interest 26 is subtracted from the estimate of the noise
signal 30 in the second adder 50, corresponding to a phase shift of 180 °. The output of the
second summer 50 is sent to the output transducer 48B and converted to an acoustic output
signal that is transmitted to the other eardrum of the user wearing the binaural hearing aid
system 10. This improves BMLD and SRT.
[0078]
FIG. 5 schematically shows a digital signal separation device 12 including an adaptive
beamformer 10 with two microphones 14, 16.
[0079]
The microphone audio signals 18, 20 are filtered in conventional prefilters 22, 24 prior to
beamforming.
The microphone audio signals 18, 20 can be digitized before or after the prefilters 22, 24 by an
A / D converter (not shown). The signals before and after prefiltering and before and after A / D
conversion are all referred to as microphone audio signals.
[0080]
The output 26 of the first subtractor 28 uses adaptive beamforming to generate an estimate of
the signal of interest from an assumed direction of interest. Then, the estimated value of the
target signal 26 is presented to one of the user's ears and presented in antiphase to the other of
the user's ears. The output 30 of the adaptive filter 32, which filters the output of the second
subtractor 34, then produces a noise estimate for presentation to the user's ears.
[0081]
The input x 1 (n) to the first microphone 14 is given by
11-04-2019
20
[0082]
[0083]
Here, h 1 (n) is an impulse response of sound propagation from a sound source emitting a signal
s (n) to the first microphone 14, and g 1 (n) is a signal q (first signal) to the first microphone 14
n) the impulse response of the speech propagation from the noise source emitting.
[0084]
The input x 2 (n) to the second microphone 16 is given by
[0085]
[0086]
Here, h 2 (n) is an impulse response of sound propagation from a sound source emitting a signal
s (n) to the second microphone 16, and g 2 (n) is a signal q (q) to the second microphone 16. n)
the impulse response of the speech propagation from the noise source emitting.
[0087]
At this time, the output 26 of the target signal is equal to h 1 (n) * s (n), and the output 30 of the
noise estimate is equal to g 1 (n) * q (n).
[0088]
FIG. 6 schematically shows a signal separation device 12 based on four microphones 14, 16, 14B
and 16B.
Among them, two microphones 14 and 16 are disposed in the first hearing aid 10A, and the
other two microphones 14B and 16B are disposed in the second hearing aid 10B.
[0089]
11-04-2019
21
The wide distance between the microphones can be used to form a directional pattern
comprising narrow beams to improve the spatial separation of the target signal estimate from the
noise estimate.
The conventional output of the beamformer may be used as the target signal estimate, or the
noise estimate may be provided by subtracting the target signal estimate from the microphone
speech signal of one of the plurality of microphones.
[0090]
The microphone sound signals 18, 20 of the two microphones 14, 16 of the first hearing aid 10A
are pre-charged to the microphone sound signals y 1 (n), y 2 (n) with respective prefilters 22, 24
known in the art. It is filtered and input to each adaptive filter a 1 (n), a 2 (n).
[0091]
The prefiltered microphone audio signals of the two microphones 14B, 16B of the second
hearing aid 10B are encoded for transmission in the second hearing aid 10B, and the first
hearing aid 10A is transmitted using wireless or wired data transmission. Sent to
Transmission data representing the microphone audio signals of the two microphones 14B, 16B
of the second hearing aid 10B is received by the transceiver 36 of the first hearing aid 10A, and
the two microphone audio signals y 3 (n), y 4 are received by the decoder 38. (N) is decoded and
input to each of the adaptive filters a 3 (n) and a 4 (n).
[0092]
Adaptive filters a 1 (n), a 2 (n), a 3 (n) and a 4 (n) are microphone sound signals y 1 (n), y 2 (n), y
3 (n) and y It is configured to filter 4 (n) to adapt each filter coefficient for adaptive beamforming
towards the source.
[0093]
Adaptive filters a 1 (n), a 2 (n), a 3 (n), a 4 (n) allow focusing on moving sound sources and while
the user of the hearing aid system is moving It is possible to concentrate on the stationary sound
11-04-2019
22
source.
Furthermore, the adaptive filters a 1 (n), a 2 (n), a 3 (n), a 4 (n) are the appearance of new sound
sources, the disappearance of noise sources or noise sources for users of hearing aid systems
Adaptable to changes in the speech environment, such as
[0094]
The adaptive beamformer filters a 1 (n), a 2 (n), a 3 (n), a 4 (n) are signals received by the at least
one microphone 14, 16, 14B, 16B, It may be designed under the assumption that it can be
modeled as a combination of a target signal from a target direction and a masker or noise.
[0095]
[0096]
Here, h i (n) is an impulse response of sound propagation from a sound source emitting a signal s
(n) to the ith microphone, and v i (n) is a noise signal in the same microphone.
The noise may consist of both directional noise and other types of noise, such as diffuse noise or
bubble noise.
[0097]
The filter coefficients are determined adaptively by solving the following optimization problem.
[0098]
[0099]
Filter adaptation is preferably performed using a least mean square (LMS) algorithm, more
preferably a normalized least mean square (NLMS) algorithm, but recursive least squares,
steepest descent or other Other algorithms may be used, such as a type of numerical
optimization algorithm.
11-04-2019
23
[0100]
The outputs of the adaptive filters a 1 (n), a 2 (n), a 3 (n), a 4 (n) are added in the adder 34, and
the output 26 of the adder 34 is an estimate of the target signal. Configure.
[0101]
[0102]
The subtractor 28 outputs an estimate of the noise.
[0103]
[0104]
Once the target signal estimate and the noise estimate are determined, the signal is presented to
the user to improve the user's SRT, as schematically shown in FIG.
[0105]
An example of a novel binaural hearing aid system 10 is shown in FIG.
[0106]
The novel binaural hearing aid system 10 comprises first and second hearing aids 10A, 10B and
comprises transceivers 36, 36B for data communication between the two hearing aids 10A, 10B.
The first hearing aid 10A includes at least one microphone with two microphones 14, 16 and
provides microphone sound signals 18, 20 in response to the sound received at each microphone
14, 16.
The microphone audio signals 18, 20 are pre-filtered into microphone audio signals at respective
prefilters 22, 24 known in the art and input to the signal separation device 12.
11-04-2019
24
The signal separation device 12 is shown in more detail in FIG. 6 and described above with
reference to FIG.
[0107]
The second hearing aid 10B also includes at least one microphone with two microphones 14B,
16B and provides microphone sound signals 18B, 20B in response to the sound received at each
microphone 14B, 16B.
The microphone audio signals 18B, 20B are prefiltered by the prefilters 22B, 24B as known in
the art.
The prefiltered microphone audio signals of the two microphones 22B, 24B are then encoded in
the codec 40B for transmission to the first hearing aid 10A using wireless data transmission.
The transmit data representing the microphone audio signal of the second hearing aid 10B is
received by the transceiver 36 of the first hearing aid 10A and input to the signal separation
device 12 at the decoder 38 as described above with reference to FIG. Into two microphone
sound signals.
[0108]
The signal separation device 12 is configured to provide an estimate of the target signal 26 and
an estimate of the noise signal 30 based on the pre-filtered microphone audio signals of the first
and second hearing aids 10A, 10B.
[0109]
Since the distance between the microphones of each hearing aid 10A, 10B is relatively large
compared to the distance between the microphones of a single hearing aid, the beamformer of
the signal separation device 12 is configured with a narrow beam directivity pattern and the
noise signal 30 It is possible to improve the spatial separation of the estimate of the signal of
interest 26 from the estimate of {circumflex over (d)} (see FIG. 6).
11-04-2019
25
The conventional output of the beamformer is used as an estimate of the target signal 26 and the
estimate of the noise signal 30 is from the prefiltered microphone speech signal of one of the
plurality or four microphones 14, 16, 14B, 16B. It is provided by subtracting the estimate of the
signal of interest 26.
[0110]
Once the target signal estimate and the noise estimate are determined, the signal is presented to
the user to improve the user's SRT.
The estimated value of the target signal 26 and the estimated value of the noise signal 30 are
added in the first adder 42, and the output sum of the estimated value of the target signal 26 and
the estimated value of the noise signal 30 is It is delayed and input to the signal processor 46 for
hearing loss compensation.
The delay unit 44 maintains the desired relative phase of the signals respectively output by the
first and second hearing aids 10A, 10B.
[0111]
An output transducer 48, which in the example shown is a receiver 48, converts the output of the
signal processor 46 into an acoustic output signal that is transmitted to the tympanic membrane
of the user wearing the binaural hearing aid system 10.
[0112]
Furthermore, the estimate of the target signal 26 is subtracted from the estimate of the noise
signal 30 in the second adder 50, corresponding to a phase shift of 180 °, and the output of the
second adder 50 is stored in the codec 40. It is encoded and transmitted by the transceiver 36 to
the second hearing aid 10B.
The transmitted sum value is received by the transceiver 36B at the second hearing aid 10B,
decoded by the decoder 38B and input to the signal processor 46B for hearing loss
11-04-2019
26
compensation.
An output transducer 48B, which in the example shown is a receiver 48B, converts the output of
the signal processor 46B into an acoustic output signal that is transmitted to the tympanic
membrane of the user wearing the binaural hearing aid system 10.
This can improve the user's SRT by up to 20 dB according to the voice environment.
[0113]
The estimate of the target signal 26 and the estimate of the noise signal 30 can be interchanged,
so instead of shifting the phase of the estimate of the target signal 26, instead of presenting the
noise signal 30 before presenting it to one of the user's eardrum. The phase of the estimate can
be shifted by 180 °.
The resulting improvement in SRT is less than that provided by the phase shift of the estimate of
the target signal 26.
11-04-2019
27
Документ
Категория
Без категории
Просмотров
0
Размер файла
39 Кб
Теги
description, jp2015039208
1/--страниц
Пожаловаться на содержимое документа