Echo cancellation method, device and storage medium

Abstract

The present disclosure relates to an echo cancellation method, device, and storage medium. The method includes: obtaining a first sound signal after linear echo cancellation of an input sound signal, and a second sound signal after linear echo residual suppression of the first sound signal; obtaining a preset nonlinear distortion signal; mapping the nonlinear distortion signal to the first sound signal to obtain a mapped signal; obtaining a second suppression value based on the first sound signal and the mapped signal; and obtaining an echo-cancelled sound signal based on the second suppression value and the second sound signal. This method can effectively eliminate nonlinear echoes in echo cancellation with low computational complexity, thereby improving the user experience.

Description

Technical Field

[0001] The present disclosure relates to the field of audio processing technology, and in particular to an echo cancellation method, device, and storage medium. Background Art

[0002] In duplex voice call scenarios, such as when a user makes a daily phone call, the far-end and near-end users communicate with each other. The far-end signal is played back through the near-end speaker. Without echo cancellation, the near-end microphone picks up the signal and mixes it with the near-end speaker's signal before transmitting it to the far-end user. This can cause the far-end user to hear their own voice. This echo can create a very poor call experience for users.

[0003] To avoid this, echo cancellation is required for duplex voice calls. Echo cancellation typically involves linear and nonlinear methods. Nonlinear echo cancellation is currently a technical challenge, with algorithms often requiring high computational complexity and ineffective cancellation. Summary of the Invention

[0004] To overcome the problems existing in the related art, the present disclosure provides an echo cancellation method, device and storage medium.

[0005] According to a first aspect of an embodiment of the present disclosure, there is provided an echo cancellation method, applied to a terminal, comprising:

[0006] Obtaining a first sound signal after linear echo cancellation is performed on an input sound signal, and a second sound signal after linear echo residual suppression is performed on the first sound signal, wherein the input sound signal includes a far-end reference signal and / or a signal collected by a microphone of the terminal, and the far-end reference signal represents a signal transmitted from the far-end to the near-end by the terminal in a duplex call scenario;

[0007] Obtaining a preset nonlinear distortion signal;

[0008] Mapping the nonlinear distortion signal to the first sound signal to obtain a mapped signal, wherein the frequency domain amplitude of the mapped signal and the first sound signal are at the same level;

[0009] obtaining a second suppression value based on the first sound signal and the mapping signal, wherein the second suppression value is used to perform nonlinear echo suppression on the input sound signal;

[0010] A sound signal after echo cancellation is obtained based on the second suppression value and the second sound signal.

[0011] In one embodiment, mapping the nonlinear distortion signal to the first sound signal includes:

[0012] mapping the nonlinear distortion signal to the first sound signal based on a mapping coefficient,

[0013] The mapping coefficient is determined when it is determined that the specified input sound signal includes a far-end reference signal but does not include a signal collected by a microphone, and the specified input sound signal includes the input sound signal or the input sound signal received most recently before the input sound signal.

[0014] In one embodiment, the second sound signal is obtained by performing linear echo residual suppression processing on the first sound signal through spectral subtraction;

[0015] The step of determining that the designated input sound signal includes the far-end reference signal but excludes the signal collected by the microphone comprises:

[0016] Determine that the voice activity detection of the far-end reference signal in the specified input sound signal is greater than a first threshold value N times in a row, where N is a positive integer; and

[0017] Determining that the voice activity detection of the second sound signal obtained according to the specified input sound signal is less than a second threshold value; and

[0018] It is determined that an average of first suppression values ​​obtained by performing spectral subtraction suppression processing on the first sound signal obtained according to the designated input sound signal is less than a third threshold value.

[0019] In one embodiment, mapping the nonlinear distortion signal to the first sound signal to obtain a mapped signal includes:

[0020] Determining a mapping coefficient using a normalized least mean square (NLMS) algorithm based on a mapping coefficient obtained from a previous mapping, a signal amplitude of the nonlinear distortion signal, a convergence factor, and an error estimate, wherein the mapping coefficient obtained from the previous mapping is a mapping coefficient determined when the input sound signal most recently received before the input sound signal includes a far-end reference signal and does not include a signal collected by a microphone;

[0021] The mapping signal is determined based on the mapping coefficients and a nonlinear distortion signal.

[0022] In one implementation, the formula for determining the mapping coefficient is expressed as follows:

[0023] W(n+1)=W(n)+mu*x(n)*e(n) / (data+x(n) 2 )

[0024] Wherein, W(n+1) is the mapping coefficient determined by this mapping, W(n) is the mapping coefficient determined by the previous mapping, n is the number of iterations, n is a positive integer, mu is the convergence factor, data is a constant greater than 0, x(n) is the amplitude of the nonlinear distortion signal before mapping, e(n) is the error estimate, e(n) = |aec_out1| - W(n) * |ref_abs|, aec_out1 is the first sound signal, ref_abs is the nonlinear distortion signal, and || is the amplitude value operation.

[0025] In one embodiment, obtaining the second suppression value based on the first sound signal and the mapping signal includes:

[0026] Calculating a signal-to-return ratio (SNR) based on the first sound signal and the mapping signal;

[0027] Performing smoothing iteration on the signal-to-return ratio SNR to obtain a smoothed iterative signal-to-return ratio SNR′;

[0028] A second suppression value is obtained based on the signal-to-return ratio SNR′ after the smoothing iteration.

[0029] In one embodiment, obtaining the second suppression value based on the signal-to-return ratio SNR′ after the smoothing iteration includes:

[0030] The second suppression value is calculated according to the formula H2=SNR' / (SNR'+1), where H2 is the second suppression value.

[0031] In one embodiment, the preset nonlinear distortion signal is obtained by performing full-wave rectification on the remote reference signal.

[0032] According to a second aspect of an embodiment of the present disclosure, there is provided an echo cancellation device, comprising:

[0033] a first acquisition module configured to acquire a first sound signal after linear echo cancellation is performed on an input sound signal, and a second sound signal after linear echo residual suppression is performed on the first sound signal, wherein the input sound signal includes a far-end reference signal and / or a signal collected by a microphone of the terminal, and the far-end reference signal represents a signal transmitted from the far-end to the near-end by the terminal in a duplex call scenario;

[0034] A second acquisition module is configured to acquire a preset nonlinear distortion signal;

[0035] a mapping module configured to map the nonlinear distortion signal to the first sound signal to obtain a mapped signal, wherein the frequency domain amplitude of the mapped signal and the first sound signal are at the same level;

[0036] a nonlinear echo cancellation module configured to obtain a second suppression value based on the first sound signal and the mapping signal, wherein the second suppression value is used to perform nonlinear echo suppression on the input sound signal; and

[0037] A sound signal after echo cancellation is obtained based on the second suppression value and the second sound signal.

[0038] In one embodiment, mapping the nonlinear distortion signal to the first sound signal includes:

[0039] mapping the nonlinear distortion signal to the first sound signal based on a mapping coefficient,

[0040] The mapping coefficient is determined when it is determined that the specified input sound signal includes a far-end reference signal but does not include a signal collected by a microphone, and the specified input sound signal includes the input sound signal or the input sound signal received most recently before the input sound signal.

[0041] In one embodiment, the second sound signal is obtained by performing linear echo residual suppression processing on the first sound signal through spectral subtraction;

[0042] The step of determining that the designated input sound signal includes the far-end reference signal but excludes the signal collected by the microphone comprises:

[0043] Determine that the voice activity detection of the far-end reference signal in the specified input sound signal is greater than a first threshold value N times in a row, where N is a positive integer; and

[0044] Determining that the voice activity detection of the second sound signal obtained according to the specified input sound signal is less than a second threshold value; and

[0045] It is determined that an average of first suppression values ​​obtained by performing spectral subtraction suppression processing on the first sound signal obtained according to the designated input sound signal is less than a third threshold value.

[0046] In one embodiment, mapping the nonlinear distortion signal to the first sound signal to obtain a mapped signal includes:

[0047] Determining a mapping coefficient using a normalized least mean square (NLMS) algorithm based on a mapping coefficient obtained from a previous mapping, a signal amplitude of the nonlinear distortion signal, a convergence factor, and an error estimate, wherein the mapping coefficient obtained from the previous mapping is a mapping coefficient determined when the input sound signal most recently received before the input sound signal includes a far-end reference signal and does not include a signal collected by a microphone;

[0048] The mapping signal is determined based on the mapping coefficients and a nonlinear distortion signal.

[0049] In one implementation, the formula for determining the mapping coefficient is expressed as follows:

[0050] W(n+1)=W(n)+mu*x(n)*e(n) / (data+x(n) 2 )

[0051] Wherein, W(n+1) is the mapping coefficient determined by this mapping, W(n) is the mapping coefficient determined by the previous mapping, n is the number of iterations, n is a positive integer, mu is the convergence factor, data is a constant greater than 0, x(n) is the amplitude of the nonlinear distortion signal before mapping, e(n) is the error estimate, e(n) = |aec_out1| - W(n) * |ref_abs|, aec_out1 is the first sound signal, ref_abs is the nonlinear distortion signal, and || is the amplitude value operation.

[0052] In one embodiment, obtaining the second suppression value based on the first sound signal and the mapping signal includes:

[0053] Calculating a signal-to-return ratio (SNR) based on the first sound signal and the mapping signal;

[0054] Performing smoothing iteration on the signal-to-return ratio SNR to obtain a smoothed iterative signal-to-return ratio SNR′;

[0055] A second suppression value is obtained based on the signal-to-return ratio SNR′ after the smoothing iteration.

[0056] In one embodiment, obtaining the second suppression value based on the signal-to-return ratio SNR′ after the smoothing iteration includes:

[0057] The second suppression value is calculated according to the formula H2=SNR' / (SNR'+1), where H2 is the second suppression value.

[0058] In one embodiment, the preset nonlinear distortion signal is obtained by performing full-wave rectification on the remote reference signal.

[0059] According to a third aspect of an embodiment of the present disclosure, there is provided an echo cancellation device, the device comprising:

[0060] processor;

[0061] a memory for storing processor-executable instructions;

[0062] The processor is configured to: execute the method as described in the first aspect or any one of the embodiments of the first aspect.

[0063] According to a fourth aspect of an embodiment of the present disclosure, a storage medium is provided, in which instructions are stored. When the instructions in the storage medium are executed by a processor of a device, the device is enabled to execute the method described in the first aspect or any one of the embodiments of the first aspect.

[0064] The technical solution provided by the embodiments of the present disclosure may include the following beneficial effects: by mapping a preset nonlinear distortion signal to a first sound signal after linear echo cancellation of an input sound signal, and then determining a suppression value for nonlinear echo suppression of the input sound signal based on the mapped signal and the first sound signal, the nonlinear echo in the echo cancellation can be better eliminated with lower computational complexity, thereby improving the user experience.

[0065] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0067] Figure 1 The figure is a schematic diagram showing echo generation in a duplex voice call scenario according to an exemplary embodiment.

[0068] Figure 2 The figure is a flowchart showing echo cancellation according to an exemplary embodiment.

[0069] Figure 3 The figure is a flow chart showing an echo cancellation method according to an exemplary embodiment.

[0070] Figure 4 The figure is a flow chart showing a method for performing linear echo cancellation on an input sound signal by using Kalman filtering according to an exemplary embodiment.

[0071] Figure 5 The figure is a flowchart of a method for performing linear echo residual suppression on a first sound signal by using spectral subtraction according to an exemplary embodiment.

[0072] FIG6(A) is a time domain and frequency domain diagram of an echo portion in a signal mic collected by a microphone according to an exemplary embodiment.

[0073] FIG6(B) is a diagram showing a remote-end reference signal in the time domain and the frequency domain according to an exemplary embodiment.

[0074] FIG6(C) is a time domain and frequency domain diagram of a nonlinear distortion signal according to an exemplary embodiment.

[0075] Figure 7 is a schematic diagram showing a nonlinear echo suppression method according to an exemplary embodiment.

[0076] Figure 8 The figure is a block diagram of an echo cancellation device according to an exemplary embodiment.

[0077] Figure 9 The figure is a block diagram showing a device for echo cancellation according to an exemplary embodiment.

[0078] Figure 10 The figure is a block diagram showing a device for echo cancellation according to an exemplary embodiment. DETAILED DESCRIPTION

[0079] Exemplary embodiments are described in detail herein, with examples illustrated in the accompanying drawings. When the following description refers to the drawings, identical numerals in different drawings represent identical or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present disclosure.

[0080] As mentioned above, in a duplex voice call scenario, echo will bring a bad call experience to users. Figure 1 FIG. 1 is a schematic diagram showing echo generation in a duplex voice call scenario according to an exemplary embodiment. Figure 1 As shown, when a far-end user is talking to a near-end user, the far-end user's microphone collects their voice signal A and transmits it to the near-end user's audio playback device. After being broadcast by the near-end user's audio playback device, voice signal A may encounter an obstacle and produce an echo A'. This echo A' may be captured by the near-end user's microphone and superimposed with the near-end user's voice signal B before being transmitted to the far-end user's audio playback device. This causes the far-end user to hear their own voice mixed in with the sound they hear, resulting in a poor user experience.

[0081] In related technologies, echo cancellation technology can be used to eliminate echoes in sound signals to improve user experience. Figure 2 FIG. 1 is a flow chart showing an echo cancellation process according to an exemplary embodiment. Figure 2 As shown, the input sound signal including the microphone signal mic and the far-end reference signal ref can be processed sequentially with linear echo cancellation, linear echo residual suppression, and nonlinear echo suppression to obtain an echo-cancelled sound signal. Here, aec_out1 is the first sound signal after linear echo cancellation, aec_out2 is the second sound signal after linear echo residual suppression, and aec_out3 is the sound signal after nonlinear echo suppression.

[0082] Figure 2In the echo cancellation process shown, nonlinear echo cancellation is currently a technical difficulty. Most of the related algorithms have high computational complexity and poor removal effect.

[0083] In view of this, an embodiment of the present disclosure provides an echo cancellation method, which maps a preset nonlinear distortion signal to a first sound signal after linear echo cancellation of an input sound signal, and then determines a suppression value for nonlinear echo suppression of the input sound signal based on the mapped signal and the first sound signal. This method can better eliminate nonlinear echoes in echo cancellation with lower computational complexity, thereby improving user experience.

[0084] Figure 3 FIG. 1 is a flow chart showing an echo cancellation method according to an exemplary embodiment. Figure 3 As shown, the echo cancellation method includes the following steps.

[0085] In step S31 , a first sound signal obtained by performing linear echo cancellation on an input sound signal and a second sound signal obtained by performing linear echo residual suppression on the first sound signal are obtained.

[0086] The input sound signal includes a signal collected by a microphone and / or a remote reference signal.

[0087] In step S32, a preset nonlinear distortion signal is obtained.

[0088] In step S33, the nonlinear distortion signal is mapped to the first sound signal to obtain a mapped signal.

[0089] The frequency domain amplitudes of the mapped signal and the first sound signal are at the same level.

[0090] In step S34 , a second suppression value is obtained based on the first sound signal and the mapping signal.

[0091] The second suppression value is used to perform nonlinear echo suppression on the input sound signal.

[0092] In step S35, a sound signal after echo cancellation is obtained based on the second suppression value and the second sound signal.

[0093] In an embodiment of the present disclosure, the input sound signal includes a far-end reference signal and / or a signal collected by a microphone of the terminal. The far-end reference signal represents the signal transmitted from the far-end to the near-end by the terminal in a duplex call scenario. The signal collected by the microphone includes the near-end sound signal and the far-end echo signal collected by the microphone. The far-end reference signal is a signal obtained by echoing the received far-end sound signal, which is the same as the sound signal sent by the far-end user to the near-end user. Linear echo cancellation is performed on the input sound signal to obtain a first sound signal. Linear echo residual suppression is performed on the first sound signal to obtain a second sound signal.

[0094] In the embodiment of the present disclosure, after obtaining the input sound signal, linear echo cancellation processing can be first performed on it to obtain a first sound signal with the linear part of the echo removed. In one example, Kalman filtering can be used to perform linear echo cancellation processing on the input sound signal. Figure 4 FIG is a flow chart showing a method for performing linear echo cancellation on an input sound signal using Kalman filtering according to an exemplary embodiment. Figure 4 As shown, voice activity detection (vad) can be performed on the far-end reference signal to determine whether there is currently a far-end speaking voice. If so, the Kalman filter coefficient is updated and filtered in combination with the signal collected by the microphone. If not, only Kalman filtering is performed. After filtering, the output result includes the first sound signal aec_out1 and the estimated echo signal echo. It is understandable that those skilled in the art may also use other methods besides Kalman filtering to achieve linear echo cancellation when implementing the technical solution of the present disclosure, and this is not limited here.

[0095] When describing the echo path, adaptive filters such as Kalman filtering used for linear echo cancellation have slow filter coefficient convergence. As the echo path changes, the filter fails to track the changes in a timely manner, resulting in residual linear echo. Therefore, it is necessary to further perform linear echo residual suppression on the first sound signal after linear echo cancellation.

[0096] In the embodiment of the present disclosure, spectral subtraction can be used to perform linear echo residual suppression on the first sound signal. The calculation formula of spectral subtraction is: Wherein, ω is the angular frequency, ω=2*π*f, H(ω) is the frequency domain suppression value, D(ω) in this embodiment can be the echo spectrum obtained after linear echo cancellation, that is, the spectrum of the echo signal echo, and Y(ω) in this embodiment can be the spectrum of the signal mic collected by the microphone.

[0097] Figure 5FIG. 1 is a flow chart showing a method for performing linear echo residual suppression on a first sound signal using spectral subtraction according to an exemplary embodiment. Figure 5 As shown, the first suppression value H1 can be calculated based on the signal mic collected by the microphone and the estimated echo signal echo, and the calculation formula is H1=1-|echo| 2 / |mic| 2 Then, the first sound signal aec_out1 is multiplied by the first suppression value H1 to obtain the second sound signal aec_out2 after the linear echo residual suppression process, that is, aec_out2 = H1*aec_out1.

[0098] Figure 6(A) is a time domain and frequency domain diagram of the echo portion of the signal mic collected by a microphone according to an exemplary embodiment, Figure 6(B) is a time domain and frequency domain diagram of a far-end reference signal according to an exemplary embodiment, and Figure 6(C) is a time domain and frequency domain diagram of a nonlinear distortion signal according to an exemplary embodiment. Figures 6(A)-6(C) As shown in the figure, the upper half is the time domain diagram of the signal, and the lower half is the frequency domain diagram of the signal. The signal in the box in Figure 6(A) is the nonlinear distortion caused by the nonlinear characteristics of the working device. This distortion is unpredictable, and as shown in Figure 6(B), since there is no signal at the corresponding position in the far-end reference signal ref, this distortion cannot be eliminated by linear echo cancellation.

[0099] In order to simplify the computational complexity of nonlinear echo suppression while satisfying the suppression effect, the embodiment of the present disclosure introduces a nonlinear distortion signal and jointly processes the nonlinear distortion signal with the first sound signal, thereby effectively eliminating the nonlinear distortion portion of the input sound signal. The nonlinear distortion signal can be introduced by full-wave rectification, that is, taking the absolute value of the far-end reference signal to obtain the introduced nonlinear distortion signal. The time domain and frequency domain diagrams of the introduced nonlinear distortion signal are shown in Figure 6 (C). It can be seen that since the nonlinear distortion signal flattens the spectrum in the frequency domain, it has a signal at any position. That is, at the position where the signal mic collected by the original microphone has a nonlinear distortion portion and the far-end reference signal has no signal, a signal is introduced in the frequency domain by the nonlinear distortion signal, so that the nonlinear distortion in the sound signal can be removed by simple filtering, thereby greatly reducing the computational complexity while ensuring the echo cancellation effect.

[0100] In the embodiment of the present disclosure, since the frequency domain amplitude of the introduced nonlinear distortion signal and the first sound signal may be quite different and subsequent filtering processing cannot be performed, it is necessary to first map the nonlinear distortion signal to the first sound signal to obtain a mapping signal, which is the mapped nonlinear distortion signal, and its frequency domain amplitude is at the same level as the frequency domain amplitude of the first sound signal. It can be understood by those skilled in the art that the frequency domain amplitudes of the two signals are at the same level, which means that the difference between the frequency domain amplitudes of the two signals is less than a preset threshold, so that the two signals can be filtered and processed. The specific value of the preset threshold can be set as needed and is not limited here.

[0101] In an embodiment of the present disclosure, after the mapping of the nonlinear distortion signal is completed, the frequency domain amplitudes of the mapped nonlinear distortion signal, i.e., the mapped signal and the first sound signal, are at the same level. At this point, the two signals can be filtered. In one example, Wiener filtering can be used to filter out the nonlinear distortion in the input sound signal. Wiener filtering can be implemented in the following manner: first, based on the first sound signal and the mapped signal, a second suppression value is obtained, and the second suppression value is used to perform nonlinear echo suppression on the input sound signal; then, based on the second suppression value and the second sound signal, a sound signal after echo cancellation is obtained.

[0102] According to the technical solution of the embodiment of the present disclosure, by mapping a preset nonlinear distortion signal to a first sound signal after linear echo cancellation of the input sound signal, and then determining the suppression value for nonlinear echo suppression of the input sound signal based on the mapped signal and the first sound signal, the nonlinear echo in the echo cancellation can be better eliminated with lower computational complexity, thereby improving the user experience.

[0103] Because frequency-domain amplitude mapping of a nonlinear distortion signal must be performed only with a far-end reference signal, otherwise new errors will be introduced, in an embodiment of the present disclosure, mapping the nonlinear distortion signal to a first sound signal includes: mapping the nonlinear distortion signal to the first sound signal based on a mapping coefficient, wherein the mapping coefficient is determined when determining that a specified input sound signal includes the far-end reference signal and does not include a signal collected by a microphone, and the specified input sound signal includes the input sound signal or the input sound signal most recently received before the input sound signal.

[0104] Furthermore, determining whether a designated input sound signal includes a far-end reference signal but excludes a signal captured by a microphone includes: determining that the voice activity detection of the far-end reference signal in the designated input sound signal is greater than a first threshold value N times in a row, where N is a positive integer; determining that the voice activity detection of the second sound signal derived from the designated input sound signal is less than a second threshold value; and determining that the average of first suppression values ​​obtained by spectral subtraction suppression processing on the first sound signal derived from the designated input sound signal is less than a third threshold value. In this manner, when all three of the above conditions are met, it can be determined that the designated input sound signal consists solely of the far-end reference signal, at which point the nonlinear distortion signal is mapped to the first sound signal to obtain a mapped signal.

[0105] It is understood that if the current input sound signal satisfies all three of the aforementioned conditions, the mapping coefficients can be updated when echo cancellation is performed on the current input sound signal, and the nonlinear distortion signal can be mapped to the first sound signal based on the updated mapping coefficients to obtain a mapped signal. Nonlinear echo suppression is then performed based on the obtained mapped signal and the first and second sound signals. If the current input sound signal does not satisfy any of the aforementioned three conditions, the mapping coefficients cannot be updated during this processing. In this case, the input sound signal that most recently satisfied the three conditions before the current input sound signal can be used as the designated sound signal. The nonlinear distortion signal can be mapped to the first sound signal based on the mapping coefficients obtained when echo cancellation was performed on the designated sound signal to obtain a mapped signal. Nonlinear echo suppression is then performed based on the obtained mapped signal and the first and second sound signals.

[0106] In an embodiment of the present disclosure, mapping the nonlinear distortion signal to the first sound signal to obtain a mapping signal can be achieved in the following manner: determining the mapping coefficient based on the mapping coefficient obtained in the previous mapping, the signal amplitude, convergence factor and error estimation of the nonlinear distortion signal through the normalized least mean square (NLMS) algorithm, wherein the mapping coefficient obtained in the previous mapping is the mapping coefficient determined when the input sound signal received most recently before the input sound signal includes the far-end reference signal but does not include the signal collected by the microphone; and determining the mapping signal based on the mapping coefficient and the nonlinear distortion signal.

[0107] Furthermore, the calculation formula of the mapping coefficient is expressed as follows: W(n+1)=W(n)+mu*x(n)*e(n) / (data+x(n) 2), where W(n+1) is the mapping coefficient determined by this mapping, W(n) is the mapping coefficient determined by the previous mapping, n is the number of iterations, n is a positive integer, mu is the convergence factor, data is a constant greater than 0, x(n) is the amplitude of the nonlinear distortion signal before mapping, x(n) = |ref_abs|, e(n) is the error estimate, e(n) = |aec_out1|-W(n)*|ref_abs|, aec_out1 is the first sound signal, ref_abs is the nonlinear distortion signal, and || is the amplitude value operation.

[0108] In an embodiment of the present disclosure, obtaining a second suppression value based on the first sound signal and the mapping signal includes: calculating a signal-to-return ratio SNR based on the first sound signal and the mapping signal; performing smoothing iteration on the signal-to-return ratio SNR to obtain a signal-to-return ratio SNR' after smoothing iteration; and obtaining a second suppression value based on the signal-to-return ratio SNR' after smoothing iteration. In one example, the second suppression value can be calculated according to the formula H2=SNR' / (SNR'+1), where H2 is the second suppression value.

[0109] Figure 7 is a schematic diagram showing a nonlinear echo suppression method according to an exemplary embodiment. Figure 7 As shown, the nonlinear echo suppression method provided by the embodiment of the present disclosure can include three main steps: nonlinear distortion introduction, frequency domain amplitude mapping, and nonlinear echo filtering. In the nonlinear distortion introduction step, in addition to the first sound signal aec_out1, the second sound signal aec_out2, the far-end reference signal ref, and the first suppression value H1 obtained by performing linear echo residual suppression on the first sound signal using spectral subtraction to obtain the second sound signal, a nonlinear distortion signal ref_abs=abs(ref) is introduced into the processed signal, where abs() represents an absolute value operation.

[0110] In the frequency-domain amplitude mapping step, as described above, it is first necessary to determine whether the input sound signal meets the condition of including the far-end reference signal and not including the signal collected by the microphone. When making the determination, it is first necessary to calculate the voice activity detection value vad1 of the reference signal ref, the voice activity detection value vad2 of the second sound signal aec_out2, and the mean value H_mean of the first suppression value H1. When vad1 > T1 (for N consecutive times), vad2 < T2, and H_mean < T3, the input sound signal meets the condition of including the far-end reference signal and not including the signal collected by the microphone, and the state where the input sound signal meets the condition of including the far-end reference signal and not including the signal collected by the microphone has been maintained for a period of time. Here, T1 is the first threshold, T2 is the second threshold, T3 is the third threshold, N is the number of determinations, N is a positive integer, and the introduction of the number of times N can ensure that during amplitude mapping, there is a transition process for amplitude transformation, thereby avoiding amplitude mutation caused by mapping at the starting position of the reference signal ref. The values of T1, T2, T3, and N can be set as needed and are not restricted here.

[0111] After determining whether the input sound signal meets the condition of including the far-end reference signal and not including the signal collected by the microphone, the mapping coefficient W from the non-linear distortion signal ref_abs to the first sound signal aec_out1 can be calculated based on NLMS. This mapping coefficient W is an iterative value, and each calculation result W(n + 1) is related to the result W(n) updated in the previous iteration. In one example, the mapping coefficient W can be calculated by the formula W(n + 1) = W(n) + mu * x(n) * e(n) / (data + x(n) 2 ), where W(n + 1) is the mapping coefficient determined in this mapping, W(n) is the mapping coefficient determined in the previous mapping, n is the number of iterations, n is a positive integer, mu is the convergence factor, data is a constant greater than 0, x(n) is the amplitude of the non-linear distortion signal before mapping, x(n) = |ref_abs|, e(n) is the error estimate, e(n) = |aec_out1| - W(n) * |ref_abs|, aec_out1 is the first sound signal, ref_abs is the non-linear distortion signal, and || is the operation of taking the amplitude value. At this time, W = W(n + 1). In another example, if the input sound signal does not meet the condition of including the far-end reference signal and not including the signal collected by the microphone, the mapping coefficient W directly takes the mapping coefficient W(n) determined in the previous mapping, and at this time, W = W(n).

[0112] After determining the mapping coefficient, the mapped signal ref_map of the non-linear distortion signal can be obtained according to the formula ref_map = ref_abs * W.

[0113] In the nonlinear echo filtering step, the nonlinear distortion echo suppression can be performed based on Wiener filtering. The Wiener filtering can include the following steps: calculating the signal-to-return ratio, that is, calculating the ratio SNR of the energy of the second sound signal to the energy of the mapped signal of the nonlinear distortion signal, SNR = |aec_out1| 2 / |ref_map| 2 ; SNR is smoothed and iterated to obtain the current signal-to-return ratio SNR' after iteration, SNR'=SNR*(1-alpha)+alpha*H2 2 *SNR_last, where alpha is the smoothing coefficient and SNR_last is the signal-to-response ratio obtained in the previous iteration. After the iteration is completed, the current signal-to-response ratio SNR' can be saved as SNR_last for the next iteration; the second suppression value H2 is calculated, H2 = SNR' / (SNR'+1).

[0114] After obtaining the second suppression value H2, the second suppression value H2 may be applied to the second sound signal, thereby obtaining an output sound signal aec_out3 after nonlinear echo suppression, aec_out3 = H2*aec_out2.

[0115] According to the technical solution of the embodiment of the present disclosure, by mapping a preset nonlinear distortion signal to a first sound signal after linear echo cancellation of the input sound signal, and then determining the suppression value for nonlinear echo suppression of the input sound signal based on the mapped signal and the first sound signal, the nonlinear echo in the echo cancellation can be better eliminated with lower computational complexity, thereby improving the user experience.

[0116] Based on the same concept, an embodiment of the present disclosure also provides an echo cancellation device.

[0117] It is understandable that the echo cancellation device provided in the embodiment of the present disclosure includes hardware structures and / or software modules corresponding to the execution of each function in order to realize the above functions. In combination with the units and algorithm steps of each example disclosed in the embodiment of the present disclosure, the embodiment of the present disclosure can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered to exceed the scope of the technical solution of the embodiment of the present disclosure.

[0118] Figure 8 FIG. 1 is a block diagram of an echo cancellation device according to an exemplary embodiment. Figure 8The device 800 includes a first acquisition module 801, a second acquisition module 802, a mapping module 803 and a nonlinear echo cancellation module 804.

[0119] The first acquisition module 801 is configured to acquire a first sound signal after linear echo cancellation is performed on an input sound signal, and a second sound signal after linear echo residual suppression is performed on the first sound signal, wherein the input sound signal includes a signal collected by a microphone and / or a far-end reference signal, and the far-end reference signal includes a picked-up signal.

[0120] The second acquisition module 802 is configured to acquire a preset nonlinear distortion signal.

[0121] The mapping module 803 is configured to map the nonlinear distortion signal to the first sound signal to obtain a mapped signal, where the frequency domain amplitude of the mapped signal and the first sound signal are at the same level.

[0122] The nonlinear echo cancellation module 804 is configured to obtain a second suppression value based on the first sound signal and the mapping signal, wherein the second suppression value is used to perform nonlinear echo suppression on the input sound signal; and

[0123] A sound signal after echo cancellation is obtained based on the second suppression value and the second sound signal.

[0124] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0125] In an embodiment of the present disclosure, mapping a nonlinear distortion signal to a first sound signal includes: mapping the nonlinear distortion signal to the first sound signal based on a mapping coefficient, wherein the mapping coefficient is determined when it is determined that a specified input sound signal includes a far-end reference signal and does not include a signal collected by a microphone, and the specified input sound signal includes the input sound signal, or the input sound signal received most recently before the input sound signal.

[0126] In an embodiment of the present disclosure, a first sound signal is subjected to linear echo residual suppression processing by spectral subtraction to obtain a second sound signal; determining whether a specified input sound signal includes a far-end reference signal but does not include a signal collected by a microphone comprises: determining that the voice activity detection of the far-end reference signal in the specified input sound signal is greater than a first threshold value N times in a row, where N is a positive integer; and determining that the voice activity detection of the second sound signal obtained according to the specified input sound signal is less than a second threshold value; and determining that the average of the first suppression values ​​obtained by spectral subtraction suppression processing on the first sound signal obtained according to the specified input sound signal is less than a third threshold value.

[0127] In an embodiment of the present disclosure, a nonlinear distortion signal is mapped to the first sound signal to obtain a mapping signal, including: determining the mapping coefficient based on the mapping coefficient obtained from the previous mapping, the signal amplitude of the nonlinear distortion signal, the convergence factor and the error estimate through the normalized least mean square (NLMS) algorithm, wherein the mapping coefficient obtained from the previous mapping is the mapping coefficient determined when the input sound signal received most recently before the input sound signal includes a far-end reference signal but does not include a signal collected by a microphone; and determining the mapping signal based on the mapping coefficient and the nonlinear distortion signal.

[0128] In the embodiment of the present disclosure, the formula for determining the mapping coefficient is expressed as follows: W(n+1)=W(n)+mu*x(n)*e(n) / (data+x(n) 2 ), where W(n+1) is the mapping coefficient determined by this mapping, W(n) is the mapping coefficient determined by the previous mapping, n is the number of iterations, n is a positive integer, mu is the convergence factor, data is a constant greater than 0, x(n) is the amplitude of the nonlinear distortion signal before mapping, e(n) is the error estimate, e(n) = |aec_out1|-W(n)*|ref_abs|, aec_out1 is the first sound signal, ref_abs is the nonlinear distortion signal, and || is the amplitude value operation.

[0129] In an embodiment of the present disclosure, obtaining a second suppression value based on the first sound signal and the mapping signal includes: calculating a signal-to-return ratio (SNR) based on the first sound signal and the mapping signal; performing smoothing iteration on the signal-to-return ratio (SNR) to obtain a signal-to-return ratio (SNR') after smoothing iteration; and obtaining the second suppression value based on the signal-to-return ratio (SNR') after smoothing iteration.

[0130] In the embodiment of the present disclosure, obtaining the second suppression value based on the smoothed iterative signal-to-return ratio SNR' includes: calculating the second suppression value according to the formula H2=SNR' / (SNR'+1), where H2 is the second suppression value.

[0131] In the embodiment of the present disclosure, the preset nonlinear distortion signal is obtained by performing full-wave rectification on the remote reference signal.

[0132] According to the technical solution of the embodiment of the present disclosure, by mapping a preset nonlinear distortion signal to a first sound signal after linear echo cancellation of the input sound signal, and then determining the suppression value for nonlinear echo suppression of the input sound signal based on the mapped signal and the first sound signal, the nonlinear echo in the echo cancellation can be better eliminated with lower computational complexity, thereby improving the user experience.

[0133] Figure 9FIG1 is a block diagram of an apparatus 900 for echo cancellation according to an exemplary embodiment. For example, the apparatus 900 may be an audio processing device, or other processing device connected to the audio processing device.

[0134] Reference Figure 9 , the apparatus 900 may include one or more of the following components: a processing component 902 , a memory 904 , a power component 906 , a multimedia component 908 , an audio component 910 , an input / output (I / O) interface 912 , a sensor component 914 , and a communication component 916 .

[0135] The processing component 902 generally controls the overall operation of the device 900, such as operations associated with display, phone calls, data communications, camera operation, and recording operations. The processing component 902 may include one or more processors 920 to execute instructions to perform all or part of the steps of the above-described method. In addition, the processing component 902 may include one or more modules to facilitate interaction between the processing component 902 and other components. For example, the processing component 902 may include a multimedia module to facilitate interaction between the multimedia component 908 and the processing component 902.

[0136] The memory 904 is configured to store various types of data to support the operations of the device 900. Examples of such data include instructions for any application or method operating on the device 900, contact data, phone book data, messages, pictures, videos, etc. The memory 904 can be implemented by any type of volatile or non-volatile storage device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0137] The power component 906 provides power to the various components of the device 900. The power component 906 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device 900.

[0138] The multimedia component 908 includes a screen that provides an output interface between the device 900 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, slides, and gestures on the touch panel. The touch sensor can not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 908 includes a front camera and / or a rear camera. When the device 900 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capabilities.

[0139] The audio component 910 is configured to output and / or input audio signals. For example, the audio component 910 includes a microphone (MIC), which is configured to receive external audio signals when the device 900 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signal can be further stored in the memory 904 or transmitted via the communication component 916. In some embodiments, the audio component 910 also includes a speaker for outputting audio signals.

[0140] I / O interface 912 provides an interface between processing component 902 and peripheral interface modules, such as a keyboard, click wheel, buttons, etc. These buttons may include but are not limited to: a home button, volume buttons, a start button, and a lock button.

[0141] The sensor assembly 914 includes one or more sensors for providing various aspects of the status assessment of the device 900. For example, the sensor assembly 914 can detect the open / closed state of the device 900, the relative positioning of components, such as the display and keypad of the device 900. The sensor assembly 914 can also detect changes in the position of the device 900 or a component of the device 900, the presence or absence of user contact with the device 900, the orientation or acceleration / deceleration of the device 900, and temperature changes of the device 900. The sensor assembly 914 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 914 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 914 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0142] The communication component 916 is configured to facilitate wired or wireless communication between the device 900 and other devices. The device 900 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 916 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 916 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0143] In an exemplary embodiment, the apparatus 900 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0144] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 904 including instructions, and the instructions can be executed by the processor 920 of the apparatus 900 to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0145] Figure 10 FIG1 is a block diagram of an apparatus 1100 for echo cancellation according to an exemplary embodiment. For example, the apparatus 1100 may be provided as a server. Figure 10 The apparatus 1100 includes a processing component 1122, which further includes one or more processors, and memory resources represented by a memory 1132 for storing instructions, such as applications, that can be executed by the processing component 1122. The applications stored in the memory 1132 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1122 is configured to execute the instructions to perform the above-described echo cancellation method.

[0146] The device 1100 may also include a power supply component 1126 configured to perform power management of the device 1100, a wired or wireless network interface 1150 configured to connect the device 1100 to a network, and an input / output (I / O) interface 1158. The device 1100 may operate based on an operating system stored in the memory 1132, such as Windows Server™, MacOS X™, Unix™, Linux™, FreeBSD™, or the like.

[0147] It is understood that in this disclosure, "plurality" refers to two or more than two, and other quantifiers are similar. "And / or" describes the association relationship of related objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects before and after are in an "or" relationship. The singular forms "a", "the" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0148] It will be further understood that the terms "first," "second," and the like are used to describe various types of information, but such information should not be limited to these terms. These terms are used solely to distinguish information of the same type from one another and do not indicate a particular order or level of importance. In fact, the terms "first," "second," and the like are fully interchangeable. For example, first information could be referred to as second information, and similarly, second information could be referred to as first information without departing from the scope of this disclosure.

[0149] It is further understood that although operations are described in a particular order in the drawings in the embodiments of the present disclosure, this should not be construed as requiring that the operations be performed in the particular order shown or in a serial order, or that all of the operations shown be performed to obtain the desired results. In certain circumstances, multitasking and parallel processing may be advantageous.

[0150] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein.

[0151] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the scope of the appended claims.

Claims

1. An echo cancellation method, characterized in that: Applied to terminals, including: Obtaining a first sound signal after linear echo cancellation is performed on an input sound signal, and a second sound signal after linear echo residual suppression is performed on the first sound signal, wherein the input sound signal includes a far-end reference signal and / or a signal collected by a microphone of the terminal, and the far-end reference signal represents a signal transmitted from the far-end to the near-end by the terminal in a duplex call scenario; Obtaining a preset nonlinear distortion signal; Mapping the nonlinear distortion signal to the first sound signal to obtain a mapped signal, wherein the frequency domain amplitudes of the mapped signal and the first sound signal are at the same level; obtaining a second suppression value based on the first sound signal and the mapping signal, wherein the second suppression value is used to perform nonlinear echo suppression on the input sound signal; A sound signal after echo cancellation is obtained based on the second suppression value and the second sound signal.

2. The method according to claim 1, characterized in that Mapping the nonlinear distortion signal to the first sound signal includes: mapping the nonlinear distortion signal to the first sound signal based on a mapping coefficient, The mapping coefficient is determined when it is determined that the specified input sound signal includes a far-end reference signal but does not include a signal collected by a microphone, and the specified input sound signal includes the input sound signal or the input sound signal received most recently before the input sound signal.

3. The method according to claim 2, characterized in that performing linear echo residual suppression processing on the first sound signal by spectral subtraction to obtain the second sound signal; The step of determining that the designated input sound signal includes the far-end reference signal but excludes the signal collected by the microphone comprises: Determining that the voice activity detection of the far-end reference signal in the designated input sound signal is greater than a first threshold value N times in a row, where N is a positive integer; and Determining that the voice activity detection of the second sound signal obtained according to the specified input sound signal is less than a second threshold value; and It is determined that an average of first suppression values ​​obtained by performing spectral subtraction suppression processing on the first sound signal obtained according to the designated input sound signal is less than a third threshold value.

4. The method according to claim 3, characterized in that Mapping the nonlinear distortion signal to the first sound signal to obtain a mapped signal includes: Determining a mapping coefficient using a normalized least mean square (NLMS) algorithm based on a mapping coefficient obtained from a previous mapping, a signal amplitude of the nonlinear distortion signal, a convergence factor, and an error estimate, wherein the mapping coefficient obtained from the previous mapping is a mapping coefficient determined when the input sound signal most recently received before the input sound signal includes a far-end reference signal and does not include a signal collected by a microphone; The mapping signal is determined based on the mapping coefficients and a nonlinear distortion signal.

5. The method according to claim 4, characterized in that The formula for determining the mapping coefficient is as follows: W(n+1)=W(n)+mu*x(n)*e(n) / (data+x(n) 2 ) Wherein, W(n+1) is the mapping coefficient determined by this mapping, W(n) is the mapping coefficient determined by the previous mapping, n is the number of iterations, n is a positive integer, mu is the convergence factor, data is a constant greater than 0, x(n) is the amplitude of the nonlinear distortion signal before mapping, e(n) is the error estimate, e(n) = |aec_out1| - W(n) * |ref_abs|, aec_out1 is the first sound signal, ref_abs is the nonlinear distortion signal, and || is the amplitude value operation.

6. The method according to claim 5, characterized in that The obtaining a second suppression value based on the first sound signal and the mapping signal includes: Calculating a signal-to-return ratio (SNR) based on the first sound signal and the mapping signal; Performing smoothing iteration on the signal-to-return ratio SNR to obtain a smoothed iterative signal-to-return ratio SNR′; A second suppression value is obtained based on the signal-to-return ratio SNR′ after the smoothing iteration.

7. The method according to claim 6, characterized in that The obtaining of a second suppression value based on the signal-to-return ratio SNR′ after the smoothing iteration includes: The second suppression value is calculated according to the formula H2=SNR' / (SNR'+1), where H2 is the second suppression value.

8. The method according to any one of claims 1 to 7, characterized in that The preset nonlinear distortion signal is obtained by performing full-wave rectification on the remote reference signal.

9. An echo cancellation device, characterized in that: include: a first acquisition module configured to acquire a first sound signal after linear echo cancellation is performed on an input sound signal, and a second sound signal after linear echo residual suppression is performed on the first sound signal, wherein the input sound signal includes a far-end reference signal and / or a signal collected by a microphone of the terminal, and the far-end reference signal represents a signal transmitted from the far-end to the near-end by the terminal in a duplex call scenario; A second acquisition module is configured to acquire a preset nonlinear distortion signal; a mapping module configured to map the nonlinear distortion signal to the first sound signal to obtain a mapped signal, wherein the frequency domain amplitude of the mapped signal and the first sound signal are at the same level; a nonlinear echo cancellation module configured to obtain a second suppression value based on the first sound signal and the mapping signal, wherein the second suppression value is used to perform nonlinear echo suppression on the input sound signal; and A sound signal after echo cancellation is obtained based on the second suppression value and the second sound signal.

10. The device according to claim 9, characterized in that Mapping the nonlinear distortion signal to the first sound signal includes: mapping the nonlinear distortion signal to the first sound signal based on a mapping coefficient, The mapping coefficient is determined when it is determined that the specified input sound signal includes a far-end reference signal but does not include a signal collected by a microphone, and the specified input sound signal includes the input sound signal or the input sound signal received most recently before the input sound signal.

11. The device according to claim 10, characterized in that performing linear echo residual suppression processing on the first sound signal by spectral subtraction to obtain the second sound signal; The step of determining that the designated input sound signal includes the far-end reference signal but excludes the signal collected by the microphone comprises: Determine that the voice activity detection of the far-end reference signal in the designated input sound signal is greater than a first threshold value N times in a row, where N is a positive integer; and Determining that the voice activity detection of the second sound signal obtained according to the specified input sound signal is less than a second threshold value; and It is determined that an average of first suppression values ​​obtained by performing spectral subtraction suppression processing on the first sound signal obtained according to the designated input sound signal is less than a third threshold value.

12. The device according to claim 11, characterized in that Mapping the nonlinear distortion signal to the first sound signal to obtain a mapped signal includes: Determining a mapping coefficient using a normalized least mean square (NLMS) algorithm based on a mapping coefficient obtained from a previous mapping, a signal amplitude of the nonlinear distortion signal, a convergence factor, and an error estimate, wherein the mapping coefficient obtained from the previous mapping is a mapping coefficient determined when the input sound signal most recently received before the input sound signal includes a far-end reference signal and does not include a signal collected by a microphone; The mapping signal is determined based on the mapping coefficients and a nonlinear distortion signal.

13. The device according to claim 12, characterized in that The formula for determining the mapping coefficient is as follows: W(n+1)=W(n)+mu*x(n)*e(n) / (data+x(n) 2 ) Wherein, W(n+1) is the mapping coefficient determined by this mapping, W(n) is the mapping coefficient determined by the previous mapping, n is the number of iterations, n is a positive integer, mu is the convergence factor, data is a constant greater than 0, x(n) is the amplitude of the nonlinear distortion signal before mapping, e(n) is the error estimate, e(n) = |aec_out1| - W(n) * |ref_abs|, aec_out1 is the first sound signal, ref_abs is the nonlinear distortion signal, and || is the amplitude value operation.

14. The device according to claim 13, characterized in that The obtaining a second suppression value based on the first sound signal and the mapping signal includes: Calculating a signal-to-return ratio (SNR) based on the first sound signal and the mapping signal; Performing smoothing iteration on the signal-to-return ratio SNR to obtain a smoothed iterative signal-to-return ratio SNR′; A second suppression value is obtained based on the signal-to-return ratio SNR′ after the smoothing iteration.

15. The device according to claim 14, characterized in that The obtaining of a second suppression value based on the signal-to-return ratio SNR′ after the smoothing iteration includes: The second suppression value is calculated according to the formula H2=SNR' / (SNR'+1), where H2 is the second suppression value.

16. The device according to any one of claims 9 to 14, characterized in that The preset nonlinear distortion signal is obtained by performing full-wave rectification on the remote reference signal.

17. An echo cancellation device, characterized in that: The device comprises: processor; a memory for storing processor-executable instructions; The processor is configured to execute the method according to any one of claims 1 to 8.

18. A storage medium, characterized in that The storage medium stores instructions, and when the instructions in the storage medium are executed by a processor of a device, the device is enabled to execute the method according to any one of claims 1 to 8.

Images