Method and system for removing howling of time domain hearing aid based on double microphones

Abstract

The invention discloses a method and a system for removing howling of a time domain hearing aid based on double microphones. The method comprises the following steps: acquiring an output signal sent by a loudspeaker device, a first signal of an internal microphone and a second signal of an external microphone; determining a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship among the output signal, the first signal and the second signal; obtaining a feedback signal after the first signal is filtered according to a preset filtering delay, the first transfer function and a second transfer function; and removing the feedback signal from the second signal to complete howling removal. According to the method and the device, the recollected feedback signal is directly obtained based on the time domain, the howling is removed by removing the feedback signal, and the calculation complexity is greatly reduced while the data processing precision is ensured.

Description

Technical Field

[0001] The present application belongs to the field of audio processing technology, and specifically relates to a method and system for removing howling from a time-domain hearing aid based on dual microphones. Background Art

[0002] Hearing aids are devices that help the hearing-impaired improve their hearing. Hearing aid whistling is a common phenomenon when wearing hearing aids. It is usually manifested as the sound signal output by the internal speaker of the hearing aid is repeatedly collected by the external microphone when it propagates in the ear canal, so that the final output sound signal is accompanied by continuous noise and murmurs, which greatly reduces the user's use effect. Therefore, how to effectively suppress whistling is crucial in the development of hearing aids, aiming to provide users with a clearer and more natural hearing experience.

[0003] Existing howling removal algorithms can be divided into two categories, one is based on frequency domain processing, and the other is based on time domain processing. Algorithms based on frequency domain processing have two disadvantages. First, in order to obtain a relatively accurate howling frequency, the frame length of the time-frequency transform needs to be longer. However, if it is too long, the complexity of the algorithm will increase sharply. If it is too short, the frequency resolution will be too low, which will affect other frequencies and thus the sound quality. Second, such algorithms must first transform the signal from the time domain to the frequency domain, and then transform the signal from the frequency domain back to the time domain after processing. These two transformations themselves bring a lot of additional computational complexity. However, although the algorithm based on time domain processing has a smaller amount of computation in principle, it also has the problem of insufficient frequency processing accuracy and easy to affect other frequencies. Summary of the invention

[0004] The present application proposes a method and system for howling removal of a time-domain hearing aid based on dual microphones, which directly obtains a re-collected feedback signal based on the time domain, and removes the howling by removing the feedback signal, thereby greatly reducing the complexity of the operation while ensuring the accuracy of data processing.

[0005] A first aspect of the present application provides a dual-microphone-based time-domain hearing aid whistle removal method, the method comprising:

[0006] collecting an output signal emitted by a loudspeaker, a first signal from an internal microphone, and a second signal from an external microphone; wherein the hearing aid comprises a loudspeaker, an internal microphone, and an external microphone;

[0007] determining a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship among the output signal, the first signal and the second signal;

[0008] Obtaining a feedback signal of the first signal after filtering according to a preset filtering delay, the first transfer function and the second transfer function;

[0009] The feedback signal is removed from the second signal to achieve anti-howling.

[0010] The above scheme collects sound signals from the speaker of the hearing aid, the internal microphone inside the ear canal, and the external microphone outside the ear canal, and is used for the subsequent calculation of the feedback signal of howling caused by the output signal emitted by the speaker being repeatedly collected by the external microphone during the propagation process in the ear canal; first, based on the propagation relationship in the ear canal, a first transfer function and a second transfer function are obtained for studying the relationship between the output signal and the first signal and the second signal; then, based on the first transfer function and the second transfer function, a feedback signal of the first signal after filtering can be obtained. After the feedback signal is collected by the external microphone, it is used as a part of the second signal. Therefore, it is necessary to delete the feedback signal from the second signal to obtain a complete external sound signal as a howling-free input signal of the speaker. After this signal is propagated by the speaker, no howling will occur due to repeated signal collection, thereby realizing high-precision hearing aid howling removal; and in the process of obtaining the feedback signal, the signal processing is based on the time domain, which does not bring additional calculation amount, greatly reducing the complexity of the calculation.

[0011] In a possible implementation method of the first aspect, the first signal and the second signal are specifically:

[0012] After the internal microphone collects the output signal, the first signal is obtained;

[0013] The external microphone collects the external sound signal and the output signal to obtain the second signal.

[0014] The above scheme shows that the first signal is the signal collected by the internal microphone when the output signal propagates inside the ear canal, and therefore can serve as data support for studying the feedback signal; the second signal includes not only the sound signal collected by the external microphone from the external environment, but also the feedback signal when the output signal propagates in the ear canal and is re-collected by the external microphone, which will cause the hearing aid to howl. Therefore, it provides data support for clearly removing the feedback signal to complete the howling removal.

[0015] In a possible implementation method of the first aspect, determining a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship between the output signal, the first signal, and the second signal is specifically:

[0016] determining a first transmission relationship between the output signal and the first signal according to a reflection principle inside the ear canal and a frequency response of a hearing aid component;

[0017] Determining a second transmission relationship between the output signal and the second signal according to an external sound signal collected by an external microphone;

[0018] Obtaining the first transfer function according to a preset historical output-internal signal sample, a preset first delay, and the first transfer relationship;

[0019] The second transfer function is obtained according to the preset historical output-external signal sample, the preset second delay and the second transfer relationship.

[0020] The above scheme first determines the relationship between the first signal and the output signal based on the influence of the hearing aid device and the reflection principle inside the ear canal; then determines the relationship between the second signal and the output signal based on the external sound signal; and then obtains the first transfer function of the changes that occur when the output signal reaches the internal microphone position and the second transfer function of the changes that occur when the output signal reaches the external microphone position based on historical data, thereby providing data support for subsequent research on how to obtain a feedback signal based on the output signal.

[0021] In a possible implementation method of the first aspect, the first transfer relationship and the second transfer relationship are specifically:

[0022] The first transfer relationship is specifically expressed as follows:

[0023] S inner =Y inner (t-D1)*H i2i ;

[0024] In the formula, S inner is the first signal, Y inner is the output signal, t is the digital domain, D1 is the first delay, H i2i is the first transfer function, * is the convolution operation;

[0025] The second transfer relationship is specifically expressed as follows:

[0026] S outer =Y inner (t-D2)*H i2o +X;

[0027] In the formula, S outer is the second signal, Y inner is the output signal, t is the digital domain, D2 is the second delay, H i2o is the second transfer function, * is the convolution operation, and X is the external sound signal.

[0028] In a possible implementation method of the first aspect, the first delay and the second delay are specifically:

[0029] The first delay is determined according to the distance between the speaker device and the internal microphone, and the system parameters of the hearing aid; the first delay is used to delay buffering the output signal;

[0030] The second delay is determined according to the distance between the speaker device and the external microphone, and the system parameters of the hearing aid; the second delay is used to delay buffering the output signal.

[0031] In the above scheme, because the hearing aid system will cause signal delay due to the audio buffer used when playing sound, in order to align the output signal with the external sound signal in time, the first delay and the second delay determined by the hearing aid system are used to buffer the output signal to improve the accuracy of signal calculation.

[0032] In a possible implementation method of the first aspect, obtaining a feedback signal after filtering the first signal according to a preset filtering delay, the first transfer function, and the second transfer function is specifically:

[0033] The difference between the second signal and the external sound signal collected by the external microphone is set as the initial feedback signal;

[0034] constructing a relationship between the initial feedback signal and the first signal according to the first transfer function and the second transfer function;

[0035] Based on the relationship between the initial feedback signal and the first signal, delay buffering the initial feedback signal according to the filtering delay to obtain a feedback signal after the first signal is filtered;

[0036] The feedback signal may cause the hearing aid to howl after being collected by the external microphone.

[0037] In a possible implementation method of the first aspect, a relationship between an initial feedback signal and the first signal is specifically:

[0038] S ref =S inner *H i2o / H i2i ;

[0039] In the formula, S ref is the initial feedback signal, S inner is the first signal, H i2i For the first transfer function, H i2o is the second transfer function, * is the convolution operation, and / is the deconvolution operation.

[0040] A second aspect of the present application provides a dual-microphone-based time-domain hearing aid whistling removal system, the system comprising: a signal acquisition module, a transfer function construction module, a feedback signal acquisition module and a hearing aid whistling removal module;

[0041] The signal acquisition module is used to acquire an output signal from a speaker, a first signal from an internal microphone, and a second signal from an external microphone; wherein the hearing aid includes a speaker, an internal microphone, and an external microphone;

[0042] The transfer function building module is used to determine a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship among the output signal, the first signal and the second signal;

[0043] The feedback signal acquisition module is used to obtain a feedback signal after the first signal is filtered according to a preset filtering delay, the first transfer function and the second transfer function;

[0044] The hearing aid howling removal module is used to remove the feedback signal from the second signal to achieve howling removal.

[0045] A third aspect of the present application provides a terminal device, the device comprising: a terminal device comprising a processor and a memory, the memory storing a computer program, and the processor implementing the steps of a dual-microphone-based time domain hearing aid whistle removal method as described in any one of the embodiments of the present application when executing the computer program.

[0046] A fourth aspect of the present application provides a storage medium storing a computer-readable program code, which, when executed, implements the steps of a dual-microphone-based time-domain hearing aid whistle removal method described in any one of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the implementation manner will be briefly introduced below. Obviously, the drawings described below are only some implementation manners of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1 It is a specific flow chart of a method for removing howling in a time domain hearing aid based on dual microphones provided in a certain embodiment of the present application;

[0049] Figure 2 This is an output signal transmission diagram of a dual-microphone-based time-domain hearing aid howling removal method provided in a certain embodiment of the present application;

[0050] Figure 3 It is a dual-microphone howling removal flow chart of a dual-microphone time-domain hearing aid howling removal method provided by a certain embodiment of the present application;

[0051] Figure 4 It is a structural diagram of a dual-microphone-based time-domain hearing aid howling removal system provided in a certain embodiment of the present application;

[0052] Figure 5 A structural diagram of a terminal device is provided for a certain embodiment of the present application. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] It should be understood that the step numbers used in this article are only for the convenience of description and are not intended to limit the order in which the steps are executed.

[0055] First embodiment

[0056] Existing methods for removing howling from hearing aids include automatically adjusting filter parameters to adapt to different listening environments and howling characteristics, processing sound signals in the frequency domain to subtract the estimated howling part to improve the quality of the sound heard, and reducing howling by suppressing the phase of the howling sound. In order to cope with various howling sounds that may occur in different environments and hearing conditions, the embodiment of the present application uses two microphones in the time domain to accurately estimate the howling part of the input signal and remove it on the basis of the above existing methods, thereby greatly reducing the complexity of howling removal while ensuring data processing accuracy.

[0057] like Figure 1 As shown, Figure 1 A specific flow chart of a method for removing howling in a time domain hearing aid based on dual microphones is provided for a certain embodiment of the present application. The method for removing howling in a time domain hearing aid based on dual microphones in this embodiment includes steps S1 to S4, which are described in detail as follows:

[0058] Step S1 , collecting an output signal from a speaker device, a first signal from an internal microphone, and a second signal from an external microphone.

[0059] In an embodiment of the present application, the hearing aid includes two microphones for collecting sound signals, one is an internal microphone located in the ear canal, and the other is an external microphone located outside the ear canal. The internal microphone collects the output signal broadcasted by the hearing aid speaker and propagated in the ear canal; the external microphone not only collects the sound signal emitted by the external environment, but also collects the output signal propagated outside the ear canal. In the embodiment of the present application, the signal collected by the internal microphone is referred to as the first signal, and the signal collected by the external microphone is referred to as the second signal.

[0060] Furthermore, the second signal is processed internally by the hearing aid, such as howling removal and noise reduction, and then is converted by D / A and played out through the speaker device, which is the output signal. However, since the output signal will propagate in the ear canal, it will be collected when it reaches the external microphone. The repeatedly collected signal is referred to as a feedback signal in the embodiment of the present application. Since the feedback signal will be superimposed on the original external sound signal to produce howling and harsh noise, the effectiveness of the hearing aid will be greatly reduced. In severe cases, the components of the hearing aid will be damaged. Therefore, it is necessary to accurately estimate the feedback signal and remove it from the second signal so that the sound signal transmitted to the speaker device is a pure external sound signal.

[0061] Therefore, the output signal emitted by the sound-speaking device is firstly collected at the speaker, and the first signal picked up by the internal microphone and the second signal picked up by the external microphone are collected.

[0062] Step S2: determining a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship among the output signal, the first signal and the second signal.

[0063] In the embodiment of the present application, in order to better demonstrate how the output signal in the hearing aid is transmitted to the two microphones, Figure 2 The output signal transmission diagram is provided. As shown in the figure, the leftmost is the speaker device, which is used to send the output signal Y inner To help users hear the outside world; at the same time, the output signal Y inner It will also be reflected in the ear canal and propagate to the internal microphone to obtain the first signal S inner , and then propagates to the external microphone to form the second signal S outer The transfer function from the speaker to the internal microphone is recorded as the first transfer function H i2i , the transfer function from the speaker to the external microphone is recorded as the second transfer function H i2o .

[0064] The first transfer function is related to the frequency response of the speaker, the reflection principle inside the ear canal and the frequency response of the internal microphone.

[0065] First, according to the output signal, the first signal and the second signal, a first transmission relationship between the output signal and the first signal, and a second transmission relationship between the output signal and the second signal are determined, as follows:

[0066] S outer =Y inner *H i2o +X;

[0067] S inner =Y inner *H i2i ;

[0068] In the formula, S inner is the first signal, Y inner is the output signal, H i2i is the first transfer function, and * is the convolution operation. outer is the second signal, Y inner is the output signal, H i2o is the second transfer function, * is the convolution operation, and X is the external sound signal.

[0069] By setting X to a specified signal, such as a frequency sweep signal, and combining the historical data corresponding to the specified signal, such as historical output-external signal samples, the second transfer function can be obtained; the first transfer function can be obtained by using the preset historical output-internal signal samples. Since different wearing methods will lead to certain calculation errors, multiple experiments are required to obtain the average value.

[0070] Illustratively, in the embodiment of the present application, the sound signal is collected in a silent environment, so X is zero, and the external microphone can only collect the feedback signal, thereby determining the second transfer function.

[0071] In addition, because the hearing aid in the embodiment of the present application first uses ping-pong buffer processing when playing sound and then obtains the output signal through the speaker device, the ping-pong buffer will cause a certain delay to the hearing aid system. This delay is related to the system parameters of the hearing aid, which is referred to as system delay here. In addition, the output signal will also produce a delay during the propagation of the ear canal. This delay is related to the distance between the speaker device and the microphone, which is referred to as propagation delay here. In order to align the output signal with the external sound signal in time and ensure the accuracy of the calculation, the output signal needs to be delayed and buffered according to the system delay and propagation delay.

[0072] Therefore, in the embodiments of the present application, the first delay D1 and the second delay D2 are set to comprehensively represent the system delay and the propagation delay of the output signal. Among them, the first delay D1 is used to delay and buffer the output signal when calculating the first transfer function; the second delay D2 is used to delay and buffer the output signal when calculating the second transfer function.

[0073] Thus, the first transfer relationship and the second transfer relationship after delay buffering can be obtained, specifically:

[0074] The first transfer relationship, the specific expression is:

[0075] S inner =Y inner (t - D1)*H i2i ;

[0076] In the formula, S inner is the first signal, Y inner is the output signal, t is the digital domain, D1 is the first delay, and H i2i is the first transfer function, and * is the convolution operation;

[0077] The second transfer relationship, the specific expression is:

[0078] S outer =Y inner (t - D2)*H i2o +X;

[0079] In the formula, S outer is the second signal, Y inner is the output signal, t is the digital domain, D2 is the second delay, H i2o is the second transfer function, * is the convolution operation, and X is the external sound signal.

[0080] As an improvement of the above solution, since the internal microphone is closer to the speaker device than the external microphone, D1 < D2 is set, and the values of D1 and D2 are related to the size of the hearing aid product.

[0081] For example, for propagation delay, if the distance between the internal microphone and the external microphone is 2 cm, then D2-D1=0.02 / 340, 340 is the speed of sound signal propagation in the air, in m / s. The specific value of system delay in D1 and D2 is determined by the number of samples to be played before the output signal is copied into the ping-pong buffer. For example, if there are 10 samples left to be played before the output signal, then the system delay is 10 samples. Assuming that the distance between the external microphone and the speaker is 3 cm, and the hearing aid sampling rate is 24 kHz, then the number of samples of propagation delay is 0.03 / 340×24000=2, that is, D2=10+2=12.

[0082] Finally, according to the first transfer relationship and the second transfer relationship after delay buffering, in combination with the historical output signal, the historical second signal and the historical first signal recorded in the experiment, the first transfer function and the second transfer function can be determined.

[0083] Step S3: obtaining a feedback signal of the first signal after filtering according to a preset filtering delay, the first transfer function and the second transfer function.

[0084] In the embodiment of the present application, the first transfer function and the second transfer function obtained in step 2 are transformed, and the difference between the second signal and the external sound signal is set as the initial feedback signal, and the following expression can be obtained:

[0085] S ref =S inner *H i2o / H i2i ;

[0086] In the formula, S ref is the initial feedback signal, S inner is the first signal, H i2i is the first transfer function, H i2o is the second transfer function, * is the convolution operation, and / is the deconvolution operation.

[0087] Because the initial feedback signal is actually the part of the second signal that re-collects the output signal, and the first signal is obtained by re-collecting the output signal by the internal microphone, so H in the above formula i2o / H i2i It is actually a transfer function of the first signal from the internal microphone to the external microphone. The above formula expresses the relationship between the initial feedback signal and the first signal.

[0088] Because H i2o and H i2iIt is obtained based on experimental statistics, so there are two problems. First, the people who wear hearing aids are different. Second, even for the same person, the way they wear hearing aids each time may be slightly different. These two solutions lead to S ref and S outer Y inner (tD)*H i2o There are errors in the part, so it is necessary to perform error checking on the second transfer function through a filter.

[0089] Optionally, in the embodiment of the present application, an FIR filter is used to filter the initial feedback signal. Figure 3 The dual-microphone anti-howling flow chart is shown. As shown in the figure, the external sound signal X is processed by the internal logic of the hearing aid and then the output signal Y is sent from the speaker device. inner , after propagating in the ear canal, the initial feedback signal S is obtained ref , the initial feedback signal is filtered through the FIR filter to obtain a higher precision feedback signal SF ref , which is the feedback signal after the first signal is filtered. Among them, SD ref It is the signal after the initial feedback signal passes through the delay module, and is the input signal of the FIR filter; SF ref is the SD after filtering ref , is the output signal of the FIR filter; "control logic" is used to control the coefficient of the FIR filter. Assuming that the coefficient of the FIR filter is h(n), the control logic is as follows:

[0090] h(n+1)=h(n)+2μX(n)SD ref (n);

[0091] Where μ is the control rate, which is a hyperparameter that determines how fast the filter follows the signal changes. Its value is usually 0.001. ref (n) is Figure 3 X and SD ref .

[0092] In addition, because the internal microphone and the external microphone receive the output signal from the speaker device through different propagation paths, the delay of the initial feedback signal is actually smaller than that of the second signal. However, the initial feedback signal will produce additional delay when passing through the FIR filter, so it is necessary to align the two signals according to the preset filter delay after the initial feedback signal and the second signal, that is, Figure 3 Medium S ref and S outer There is also a delay module afterwards.

[0093] Step S4: removing the feedback signal from the second signal to achieve howling removal.

[0094] In the embodiment of the present application, the main reason for the hearing aid to produce howling is that the external microphone collects the feedback signal, and the feedback signal is amplified by the internal logic of the hearing aid to form positive feedback. Therefore, by removing the feedback signal from the second signal, the hearing aid can be howled and a pure external sound signal can be obtained.

[0095] Implementing the embodiments of the present application has the following beneficial effects:

[0096] The embodiment of the present application collects sound signals from the speaker of the hearing aid, the internal microphone inside the ear canal, and the external microphone outside the ear canal, so as to subsequently calculate the feedback signal of howling caused by the output signal emitted by the speaker being repeatedly collected by the external microphone during the propagation process in the ear canal; first, according to the propagation relationship in the ear canal, a first transfer function and a second transfer function are obtained for studying the relationship between the output signal and the first signal and the second signal; then, according to the first transfer function and the second transfer function, a feedback signal of the first signal after filtering can be obtained. After being collected by the external microphone as part of the second signal, the feedback signal will be superimposed with the external sound signal to produce howling. Therefore, it is necessary to remove the feedback signal from the second signal to obtain a complete external sound signal as a pure input signal of the speaker. After this signal is propagated by the speaker, no howling will occur due to the positive feedback of the signal, thereby realizing high-precision hearing aid howling removal; and in the process of obtaining the feedback signal, the processing of the signal is based on the time domain, which does not bring additional calculation amount, greatly reducing the complexity of the calculation. In addition, we also take into account that the historical data for calculating the transfer function has certain errors due to the different people and methods of wearing hearing aids, so the feedback signal is delayed and buffered through an adaptive filter to align the feedback signal and the second signal with each other, further improving the data accuracy.

[0097] Second embodiment

[0098] Further, in order to implement the dual-microphone-based time-domain hearing aid whistling removal system corresponding to the above method embodiment to achieve corresponding functions and technical effects, Figure 4 A structural diagram of a dual-microphone-based time-domain hearing aid whistle removal system is provided. For ease of description, only the parts related to this embodiment are shown. The dual-microphone-based time-domain hearing aid whistle removal system provided in the embodiment of the present application includes:

[0099] The signal acquisition module 201 is used to acquire the output signal from the speaker, the first signal from the internal microphone and the second signal from the external microphone; wherein the hearing aid includes the speaker, the internal microphone and the external microphone.

[0100] In an embodiment of the present application, the hearing aid includes two microphones for collecting sound signals, one is an internal microphone located in the ear canal, and the other is an external microphone located outside the ear canal. The internal microphone collects the output signal broadcasted by the hearing aid speaker and propagated in the ear canal; the external microphone not only collects the sound signal emitted by the external environment, but also collects the output signal propagated outside the ear canal. In the embodiment of the present application, the signal collected by the internal microphone is referred to as the first signal, and the signal collected by the external microphone is referred to as the second signal.

[0101] Furthermore, the second signal is processed internally by the hearing aid, such as howling removal and noise reduction, and then is converted by D / A and played out through the speaker device, which is the output signal. However, since the output signal will propagate in the ear canal, it will be collected when it reaches the external microphone. The repeatedly collected signal is referred to as a feedback signal in the embodiment of the present application. Since the feedback signal will be superimposed on the original external sound signal to produce howling and harsh noise, the effectiveness of the hearing aid will be greatly reduced. In severe cases, the components of the hearing aid will be damaged. Therefore, it is necessary to accurately estimate the feedback signal and remove it from the second signal so that the sound signal transmitted to the speaker device is a pure external sound signal.

[0102] The transfer function building module 202 is used to determine a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship among the output signal, the first signal and the second signal.

[0103] In an embodiment of the present application, a first transfer relationship between the output signal and the first signal is determined based on the reflection principle inside the ear canal and the frequency response of the hearing aid element; a second transfer relationship between the output signal and the second signal is determined based on the external sound signal collected by the external microphone; the first transfer function is obtained based on a preset historical output-internal signal sample, a preset first delay and the first transfer relationship; and the second transfer function is obtained based on a preset historical output-external signal sample, a preset second delay and the second transfer relationship.

[0104] The feedback signal acquisition module 203 is used to obtain a feedback signal after the first signal is filtered according to a preset filtering delay, the first transfer function and the second transfer function.

[0105] In the embodiment of the present application, the difference between the second signal and the external sound signal collected by the external microphone is set as the initial feedback signal; according to the first transfer function and the second transfer function, the relationship between the initial feedback signal and the first signal is constructed; based on the relationship between the initial feedback signal and the first signal, the initial feedback signal is delayed and buffered according to the filtering delay to obtain a feedback signal after the first signal is filtered;

[0106] The feedback signal may cause the hearing aid to howl after being collected by the external microphone.

[0107] The hearing aid howling removal module 204 is configured to remove the feedback signal from the second signal to achieve howling removal.

[0108] In the embodiment of the present application, the main reason for the hearing aid to produce howling is that the external microphone collects the feedback signal, and the feedback signal is amplified by the internal logic of the hearing aid to form positive feedback. Therefore, by removing the feedback signal from the second signal, the hearing aid can be howled and a pure external sound signal can be obtained.

[0109] In some specific embodiments, the transfer function building module 202 further includes:

[0110] First, according to the output signal, the first signal and the second signal, a first transmission relationship between the output signal and the first signal, and a second transmission relationship between the output signal and the second signal are determined, as follows:

[0111] S outer =Y inner *H i2o +X;

[0112] S inner =Y inner *H i2i ;

[0113] In the formula, S inner is the first signal, Y inner is the output signal, H i2i is the first transfer function, and * is the convolution operation. outer is the second signal, Y inner is the output signal, H i2o is the second transfer function, * is the convolution operation, and X is the external sound signal.

[0114] The first transfer function is related to the frequency response of the speaker, the reflection principle inside the ear canal and the frequency response of the internal microphone.

[0115] By setting X to a specified signal, such as a frequency sweep signal, and combining the historical data corresponding to the specified signal, such as historical output-external signal samples, the second transfer function can be obtained; the first transfer function can be obtained by using the preset historical output-internal signal samples. Since different wearing methods will lead to certain calculation errors, multiple experiments are required to obtain the average value.

[0116] Illustratively, in the embodiment of the present application, the sound signal is collected in a silent environment, so X is zero, and the external microphone can only collect the feedback signal, thereby determining the second transfer function.

[0117] In addition, because the hearing aid in the embodiment of the present application first uses ping-pong buffer processing when playing sound and then obtains the output signal through the speaker device, the ping-pong buffer will cause a certain delay to the hearing aid system. This delay is related to the system parameters of the hearing aid, which is referred to as system delay here. In addition, the output signal will also produce a delay during propagation in the ear canal. This delay is related to the distance between the speaker device and the microphone, which is referred to as propagation delay here. In order to align the output signal with the external sound signal in time and ensure the accuracy of the calculation, the output signal needs to be delayed and buffered according to the system delay and propagation delay.

[0118] Therefore, in the embodiment of the present application, a first delay D1 and a second delay D2 are set to comprehensively represent the system delay and propagation delay of the output signal. The first delay D1 is used to delay buffer the output signal when calculating the first transfer function; the second delay D2 is used to delay buffer the output signal when calculating the second transfer function.

[0119] Thus, the first transmission relationship and the second transmission relationship after delay buffering can be obtained, which are specifically:

[0120] The first transfer relationship is specifically expressed as follows:

[0121] S inner =Y inner (t-D1)*H i2i ;

[0122] In the formula, S inner is the first signal, Y inner is the output signal, t is the digital domain, D1 is the first delay, H i2i is the first transfer function, * is the convolution operation;

[0123] The second transfer relationship is specifically expressed as follows:

[0124] S outer =Y inner (t-D2)*H i2o +X;

[0125] In the formula, S outer is the second signal, Y inner is the output signal, t is the digital domain, D2 is the second delay, H i2o is the second transfer function, * is the convolution operation, and X is the external sound signal.

[0126] As an improvement to the above solution, since the internal microphone is closer to the speaker device than the external microphone, D1 < D2 is set, and the values of D1 and D2 are related to the size of the hearing aid product.

[0127] Exemplarily, for the propagation delay, if the distance between the internal microphone and the external microphone is 2 cm, then D2 - D1 = 0.02 / 340, where 340 is the speed of sound signal propagation in air, with the unit of m / s. The specific values of the system delay in D1 and D2 are determined by the number of samples to be played in front of the output signal when copying the output signal into the ping-pong buffer. For example, if there are 10 samples left to be played in front of the output signal, then the system delay is 10 samples. Assuming the distance between the external microphone and the speaker device is 3 cm and the hearing aid sampling rate is 24 kHz, then the number of samples of the propagation delay is 0.03 / 340 × 24000 = 2, that is, D2 = 10 + 2 = 12.

[0128] Finally, according to the first transfer relationship and the second transfer relationship after delay buffering, and in combination with the historical output signal, historical second signal, and historical first signal recorded in the experiment, the first transfer function and the second transfer function can be determined.

[0129] In some specific embodiments, the feedback signal acquisition module 203 further includes:

[0130] By transforming the first transfer function and the second transfer function, and making the difference between the second signal and the external sound signal the initial feedback signal, the following expression can be obtained:

[0131] S ref = S inner * H i2o / H i2i ;

[0132] In the formula, S ref is the initial feedback signal, S inner is the first signal, H i2i is the first transfer function, H i2o is the second transfer function, * is the convolution operation, and / is the deconvolution operation.

[0133] Because the initial feedback signal is actually the part of the second signal that re-collects the output signal, and the first signal is obtained by re-collecting the output signal by the internal microphone, so H in the above formula i2o / H i2i It is actually a transfer function of the first signal from the internal microphone to the external microphone. The above formula expresses the relationship between the initial feedback signal and the first signal.

[0134] Because H i2o and H i2i It is obtained based on experimental statistics, so there are two problems. First, the people who wear hearing aids are different. Second, even for the same person, the way they wear hearing aids each time may be slightly different. These two solutions lead to S ref and S outer Y inner (tD)*H i2o There are errors in the part, so it is necessary to perform error checking on the second transfer function through a filter.

[0135] Optionally, in the embodiment of the present application, an FIR filter is used to filter the initial feedback signal. The specific filtering process is as follows: the external sound signal X is processed by the internal logic of the hearing aid and then an output signal Y is sent from the speaker device. inner , after propagating in the ear canal, the initial feedback signal S is obtained ref , the initial feedback signal is filtered through the FIR filter to obtain a higher precision feedback signal SF ref , which is the feedback signal after the first signal is filtered.

[0136] The FIR filter further includes a control module for controlling the coefficient of the FIR filter. Assuming that the coefficient of the FIR filter is h(n), the control logic is as follows:

[0137] h(n+1)=h(n)+2μX(n)SD ref (n);

[0138] In the formula, μ is the control rate, which is a hyperparameter that determines how fast the filter follows the signal changes, and its value is generally 0.001.

[0139] In addition, because the internal microphone and the external microphone receive the output signal sent from the speaker device through different propagation paths, the delay of the initial feedback signal is actually smaller than that of the second signal. However, the initial feedback signal will produce additional delay when passing through the FIR filter, so it is necessary to align the two signals according to the preset filter delay after the initial feedback signal and the second signal.

[0140] Implementing the embodiments of the present application has the following beneficial effects:

[0141] The embodiment of the present application collects sound signals from the speaker of the hearing aid, the internal microphone inside the ear canal, and the external microphone outside the ear canal, so as to subsequently calculate the feedback signal of howling caused by the output signal emitted by the speaker being repeatedly collected by the external microphone during the propagation process in the ear canal; first, according to the propagation relationship in the ear canal, a first transfer function and a second transfer function are obtained for studying the relationship between the output signal and the first signal and the second signal; then, according to the first transfer function and the second transfer function, a feedback signal of the first signal after filtering can be obtained. After being collected by the external microphone as part of the second signal, the feedback signal will be superimposed with the external sound signal to produce howling. Therefore, it is necessary to remove the feedback signal from the second signal to obtain a complete external sound signal as a pure input signal of the speaker. After this signal is propagated by the speaker, no howling will occur due to the positive feedback of the signal, thereby realizing high-precision hearing aid howling removal; and in the process of obtaining the feedback signal, the processing of the signal is based on the time domain, which does not bring additional calculation amount, greatly reducing the complexity of the calculation. In addition, we also take into account that the historical data for calculating the transfer function has certain errors due to the different people and methods of wearing hearing aids, so the feedback signal is delayed and buffered through an adaptive filter to align the feedback signal and the second signal with each other, further improving the data accuracy.

[0142] Furthermore, Figure 5 This is a structural diagram of a terminal device provided in one embodiment of the present application. Figure 5 As shown, the terminal device 3 of this embodiment includes: at least one processor 30 (in Figure 5 Only one is shown) and a memory 31 and a computer program 32 stored in the memory 31 and executable on the at least one processor. When the processor 30 executes the computer program 32, the steps of a method for removing howling in a time domain hearing aid based on dual microphones as described in any one of the embodiments of the present application can be implemented.

[0143] The terminal device 3 may be a computing device such as a desktop computer, a cloud server, or a laptop computer. The computing device may include but is not limited to a processor 30 and a memory 31 . Figure 5 This is merely an example of the terminal device 3 and does not constitute a limitation on the terminal device 3 , which may include more or fewer components than those shown in the figure.

[0144] An embodiment of the present application provides a storage medium storing a computer-readable program code. When the computer-readable program code is executed, the steps of the above-mentioned method for removing howling in a time-domain hearing aid based on dual microphones are implemented.

[0145] The specific embodiments described above further describe the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the scope of protection of the present application. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A dual-microphone time-domain hearing aid howling removal method, characterized in that: include: collecting an output signal emitted by a loudspeaker, a first signal from an internal microphone, and a second signal from an external microphone; wherein the hearing aid comprises a loudspeaker, an internal microphone, and an external microphone; determining a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship among the output signal, the first signal and the second signal; Obtaining a feedback signal of the first signal after filtering according to a preset filtering delay, the first transfer function and the second transfer function; The feedback signal is removed from the second signal to achieve anti-howling.

2. The method for removing howling from a time-domain hearing aid based on dual microphones according to claim 1, characterized in that: The first signal and the second signal are specifically: After the internal microphone collects the output signal, the first signal is obtained; The external microphone collects the external sound signal and the output signal to obtain the second signal.

3. The method for removing howling from a time-domain hearing aid based on dual microphones according to claim 1, characterized in that: The determining, according to the transfer relationship between the output signal, the first signal and the second signal, a first transfer function of the first signal and a second transfer function of the second signal is specifically: determining a first transmission relationship between the output signal and the first signal according to a reflection principle inside the ear canal and a frequency response of a hearing aid component; Determining a second transmission relationship between the output signal and the second signal according to an external sound signal collected by an external microphone; Obtaining the first transfer function according to a preset historical output-internal signal sample, a preset first delay, and the first transfer relationship; The second transfer function is obtained according to the preset historical output-external signal sample, the preset second delay and the second transfer relationship.

4. The method for removing howling from a time-domain hearing aid based on dual microphones according to claim 3, characterized in that: The first transfer relationship and the second transfer relationship are specifically: The first transfer relationship is specifically expressed as follows: S inner =Y inner (t-D1)*H i2i ; In the formula, S inner is the first signal, Y inner is the output signal, t is the digital domain, D1 is the first delay, H i2i is the first transfer function, * is the convolution operation; The second transfer relationship is specifically expressed as follows: S outer =Y inner (t-D2)*H i2o +X; In the formula, S outer is the second signal, Y inner is the output signal, t is the digital domain, D2 is the second delay, H i2o is the second transfer function, * is the convolution operation, and X is the external sound signal.

5. The method for removing howling from a time-domain hearing aid based on dual microphones according to claim 3, characterized in that: The first delay and the second delay are specifically: The first delay is determined according to the distance between the speaker device and the internal microphone, and the system parameters of the hearing aid; the first delay is used to delay buffering the output signal; The second delay is determined according to the distance between the speaker device and the external microphone, and the system parameters of the hearing aid; the second delay is used to delay buffering the output signal.

6. The method for removing howling from a time-domain hearing aid based on dual microphones according to claim 1, characterized in that: The step of obtaining a feedback signal after filtering the first signal according to the preset filtering delay, the first transfer function and the second transfer function is specifically: Let the difference between the second signal and the external sound signal collected by the external microphone be the initial feedback signal; constructing a relationship between the initial feedback signal and the first signal according to the first transfer function and the second transfer function; Based on the relationship between the initial feedback signal and the first signal, delay buffering the initial feedback signal according to the filtering delay to obtain a feedback signal after the first signal is filtered; The feedback signal may cause the hearing aid to howl after being collected by the external microphone.

7. The method for removing howling from a time-domain hearing aid based on dual microphones according to claim 6, characterized in that: The relationship between the initial feedback signal and the first signal is specifically: S ref =S inner *H i2o / H i2i ; In the formula, S ref is the initial feedback signal, S inner is the first signal, H i2i For the first transfer function, H i2o is the second transfer function, * is the convolution operation, and / is the deconvolution operation.

8. A dual-microphone time-domain hearing aid howling removal system, characterized in that: include: Signal acquisition module, transfer function building module, feedback signal acquisition module and hearing aid anti-howling module; The signal acquisition module is used to acquire an output signal from a speaker, a first signal from an internal microphone, and a second signal from an external microphone; wherein the hearing aid includes a speaker, an internal microphone, and an external microphone; The transfer function building module is used to determine a first transfer function of the first signal and a second transfer function of the second signal according to a transfer relationship among the output signal, the first signal and the second signal; The feedback signal acquisition module is used to obtain a feedback signal after the first signal is filtered according to a preset filtering delay, the first transfer function and the second transfer function; The hearing aid howling removal module is used to remove the feedback signal from the second signal to achieve howling removal.

9. A terminal device, characterized in that: The invention comprises a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a method for removing howling in a time domain hearing aid based on dual microphones according to any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: The storage medium stores computer-readable program codes, and when the computer-readable program codes are executed, the steps of the method for removing howling in a time-domain hearing aid based on dual microphones according to any one of claims 1 to 7 are implemented.

Images