CN109727605B - Method and system for processing sound signal - Google Patents

Abstract

The invention discloses a method and a system for processing a sound signal. One embodiment of the method comprises: acquiring a sound signal to be processed, wherein the sound signal to be processed comprises a target sound signal and an interference sound signal; determining the power spectral density of the interference sound signal, and performing weighting processing on the sound signal to be processed according to the power spectral density to obtain a frequency spectrum estimation of a target sound signal; determining a masking threshold from the spectral estimate; and under the condition that the frequency spectrum component of the interference sound signal in the sound signal to be processed is determined to be larger than the masking threshold, carrying out filtering processing on the sound signal to be processed. The method can reduce the distortion of the sound signal, sound more naturally, reduce the complexity of algorithm calculation and accelerate the convergence speed of the preposed echo canceller. And the robustness under strong background noise and near-end voice environment can be improved.

Description

Method and system for processing sound signals

technical field

The present invention relates to the technical field of signal processing, and in particular, to a method and system for processing sound signals.

Background technique

In the prior art, the filtering processing of the sound signal can reduce "music noise", but there is a problem that the speech signal after the noise reduction processing by the filter is unnatural to a certain extent. Because the human ear is likely to be interfered and suppressed by another sound when it receives a sound, this phenomenon is called the masking effect. The closer the pitch or time of the two sounds, the more serious the masking effect, so the residual noise after noise reduction processing by the post-filter generally loses its original characteristics, which makes the hearing test unnatural to a certain extent.

SUMMARY OF THE INVENTION

Embodiments of the present invention provide a method and system for processing a sound signal, which are used to solve at least one of the above technical problems.

In a first aspect, an embodiment of the present invention provides a method for processing a sound signal, including: acquiring a sound signal to be processed, where the sound signal to be processed includes a target sound signal and an interference sound signal; and determining a power spectral density of the interference sound signal , and perform weighting processing on the sound signal to be processed according to the power spectral density to obtain a spectral estimation of the target sound signal; determine a masking threshold according to the spectral estimation; determine the frequency spectrum of the interfering sound signal in the sound signal to be processed When the component is greater than the masking threshold, filter processing is performed on the to-be-processed sound signal.

Optionally, the interfering sound signal includes a noise signal and an echo signal.

Optionally, the step of performing weighting processing on the to-be-processed sound signal according to the power spectral density to obtain the spectral estimation of the target sound signal includes:

Convert the to-be-processed sound signal into a frequency domain signal E(Ω);

Determine the posterior signal-to-noise ratio PostSNR(Ω) according to the following formula:

PostSNR(Ω)=|E(Ω)| ² /(R _bb (Ω)+R _nn (Ω)),

Wherein, R _bb (Ω) is the power spectral density of the echo signal, and R _nn (Ω) is the power spectral density of the noise signal;

The priori SNR(Ω) is derived from the following formula:

PrioriSNR(Ω _i )=(1-alpha)*P(PostSNR(Ω _i )-1)+alpha*|S'(Ω _i-1 )| ² /R _bb (Ω);

Among them, alpha is the smoothing factor, P(x)=(|x|+x)/2, S'(Ω _i-1 ) is the spectral estimation of the sound signal of the previous frame;

Further calculate the weighting coefficient H _LSA (Ω), and obtain the spectral estimation S' (Ω) of the target sound signal:

S'(Ω)=E(Ω)*H _LSA (Ω),

Wherein, theta=PostSNR(Ω)*PrioriSNR(Ω)/(PrioriSNR(Ω)+1).

Optionally, when it is determined that the spectral component of the interfering sound signal in the sound signal to be processed is greater than the masking threshold, the step of filtering the sound signal to be processed includes:

According to the power spectral density of the echo signal and the power spectral density of the noise signal, the weighting coefficient H(Ω) of the filtering process is determined:

H(Ω)=min(1, sqrt(R _TT (Ω)/(R _bb (Ω)+R _nn (Ω))) +(zeta_b*R _bb (Ω)+zeta_n*R _nn (Ω))/ (R _bb (Ω)+R _nn (Ω))),

Wherein, R _bb (Ω) is the power spectral density of the echo signal, R _nn (Ω) is the power spectral density of the noise signal, zeta_b is the echo attenuation coefficient, and zeta_n is the noise attenuation coefficient.

Optionally, the step of determining a masking threshold according to the spectrum estimation includes:

According to the spectrum estimation, determine the power spectral density B(k) of the critical band and the extended critical band spectrum C(k) of the sound signal to be processed:

C(k)=B(k)*SF(k),

Among them, SF(k)=15.81+7.5*k+0.474-17.5*sqrt(1+(k+0.474)2), bh, bl are the upper and lower limit frequencies of each critical band respectively;

According to the extended critical band spectrum C(k) and the offset function O(k), determine the preliminary masking threshold T(k):

T(k)=10 ^{lg(C(k))-(O(k)/10)} ,

Among them, the offset function O(k)=belta*(14.5+k)+(1-belta)*5.5; belta is the pitch coefficient;

From the preliminary masking threshold T(k) and the absolute hearing threshold T _abs (k), determine the masking threshold R _TT (Ω):

R _TT (Ω)=min(T(k),T _abs (k)),

Wherein, T _abs (k)=3.64f ^-0.8 -6.5exp(f-3.3) ² +10 ^-3 f ⁴ .

Optionally, the step of acquiring the sound signal to be processed includes:

receive the initial sound signal;

Perform echo cancellation on the initial sound signal to obtain the to-be-processed sound signal.

Optionally, the sound signal to be processed is a voice signal.

In a second aspect, an embodiment of the present invention provides a system for processing sound signals, including: a signal acquisition module for acquiring a to-be-processed sound signal, where the to-be-processed sound signal includes a target sound signal and an interference sound signal; a spectrum estimation and determination module , for determining the power spectral density of the interfering sound signal, and weighting the to-be-processed sound signal according to the power spectral density to obtain a spectral estimation of the target sound signal; a masking threshold determination module for The spectral estimation determines a masking threshold; a filtering processing module is configured to filter the to-be-processed sound signal when it is determined that the spectral component of the interfering sound signal in the to-be-processed sound signal is greater than the masking threshold.

Optionally, the interfering sound signal includes a noise signal and an echo signal.

Optionally, the spectrum estimation and determination module is further configured to convert the to-be-processed sound signal into a frequency domain signal E (Ω); and, determine the posterior signal-to-noise ratio PostSNR (Ω) according to the following formula:

PostSNR(Ω)=|E(Ω)| ² /(R _bb (Ω)+R _nn (Ω)),

Wherein, R _bb (Ω) is the power spectral density of the echo signal, and R _nn (Ω) is the power spectral density of the noise signal;

The priori SNR(Ω) is derived from the following formula:

PrioriSNR(Ω _i )=(1-alpha)*P(PostSNR(Ω _i )-1)+alpha*|S'(Ω _i-1 )| ² /R _bb (Ω);

Among them, alpha is the smoothing factor, P(x)=(|x|+x)/2, S'(Ω _i-1 ) is the spectral estimation of the sound signal of the previous frame;

Further calculate the weighting coefficient H _LSA (Ω), and obtain the spectral estimation S' (Ω) of the target sound signal:

S'(Ω)=E(Ω)*H _LSA (Ω),

Wherein, theta=PostSNR(Ω)*PrioriSNR(Ω)/(PrioriSNR(Ω)+1).

Optionally, the masking threshold determination module is further configured to, according to the spectrum estimation, determine the power spectral density B(k) and the extended critical band spectrum C(k) of the critical band of the sound signal to be processed:

C(k)=B(k)*SF(k),

Among them, SF(k)=15.81+7.5*k+0.474-17.5*sqrt(1+(k+0.474)2), bh, bl are the upper and lower limit frequencies of each critical band respectively;

According to the extended critical band spectrum C(k) and the offset function O(k), determine the preliminary masking threshold T(k):

T(k)=10 ^{lg(C(k))-(O(k)/10)} ,

Among them, the offset function O(k)=belta*(14.5+k)+(1-belta)*5.5; belta is the pitch coefficient;

From the preliminary masking threshold T(k) and the absolute hearing threshold T _abs (k), determine the masking threshold R _TT (Ω):

R _TT (Ω)=min(T(k),T _abs (k)),

Wherein, T _abs (k)=3.64f ^-0.8 -6.5exp(f-3.3) ² +10 ^-3 f ⁴ .

Optionally, the filtering processing module is further configured to determine the weighting coefficient H (Ω) of filtering processing according to the power spectral density of the echo signal and the power spectral density of the noise signal:

H(Ω)=min(1, sqrt(R _TT (Ω)/(R _bb (Ω)+R _nn (Ω))) +(zeta_b*R _bb (Ω)+zeta_n*R _nn (Ω))/ (R _bb (Ω)+R _nn (Ω))),

Wherein, R _bb (Ω) is the power spectral density of the echo signal, R _nn (Ω) is the power spectral density of the noise signal, zeta_b is the echo attenuation coefficient, and zeta_n is the noise attenuation coefficient.

Optionally, the signal acquisition module is further configured to receive an initial sound signal; perform echo cancellation on the initial sound signal to obtain the to-be-processed sound signal.

In a third aspect, an embodiment of the present invention provides a storage medium, where one or more programs including execution instructions are stored in the storage medium, and the execution instructions can be used by an electronic device (including but not limited to a computer, a server, or a network). device, etc.) to read and execute, so as to execute any one of the above-mentioned methods for processing sound signals of the present invention.

In a fourth aspect, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, The instructions are executed by the at least one processor to enable the at least one processor to perform any of the above-described methods and systems for processing sound signals of the present invention.

In a fifth aspect, an embodiment of the present invention further provides a computer program product, the computer program product includes a computer program stored on a storage medium, the computer program includes program instructions, and when the program instructions are executed by a computer, causes the The computer executes any of the above-mentioned methods and systems for processing sound signals.

The beneficial effect of the embodiment of the present invention is that the distortion of the sound signal can be reduced, and the sound is more natural, and the masking threshold is further determined by calculating the power spectral density PSD of the interference sound signal, which reduces the complexity of algorithm calculation. And it reduces the order requirement of the pre-echo cancel filter, thereby accelerating the convergence speed of the pre-echo canceler. And, it can improve its robustness in strong background noise and near-end speech environment.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings used in the description of the embodiments. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained from these drawings without any creative effort.

FIG. 1 is a flowchart of an embodiment of a method for processing a sound signal of the present invention;

2 is a flowchart of another embodiment of the method for processing a sound signal of the present invention;

3 is a schematic diagram of an embodiment of a system for implementing a method for processing a speech signal according to the present invention;

FIG. 4 is a schematic diagram of an embodiment of a method for processing a speech signal of the present invention;

FIG. 5 is a schematic diagram of an embodiment of a system for processing sound signals of the present invention;

FIG. 6 is a schematic structural diagram of an embodiment of an electronic device of the present invention.

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

It should be noted that the embodiments in the present application and the features of the embodiments may be combined with each other under the condition of no conflict.

The invention may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, elements, data structures, etc. that perform particular tasks or implement particular abstract data types. The invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including storage devices.

In the present invention, "module", "device", "system", etc. refer to relevant entities applied to a computer, such as hardware, a combination of hardware and software, software or software in execution, and the like. In detail, for example, an element can be, but is not limited to, a process running on a processor, a processor, an object, an executable element, a thread of execution, a program, and/or a computer. Also, an application or script running on a server, the server can be a component. One or more elements may be in a process and/or thread of execution and an element may be localized on one computer and/or distributed between two or more computers and may be executed from various computer readable media . Elements may also pass through a signal having one or more data packets, for example, a signal from one interacting with another element in a local system, in a distributed system, and/or with data interacting with other systems through a network of the Internet local and/or remote processes to communicate.

Finally, it should also be noted that in this document, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply these entities or that there is any such actual relationship or sequence between operations. Moreover, the terms "comprising" and "comprising" include not only those elements, but also other elements not expressly listed, or elements inherent to such a process, method, article or apparatus. Without further limitation, an element defined by the phrase "comprising" does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

As shown in FIG. 1, an embodiment of the present invention provides a method for processing a sound signal, including:

Step S11: Acquire a to-be-processed sound signal, where the to-be-processed sound signal includes a target sound signal and an interference sound signal.

Step S12: Determine the power spectral density of the interfering sound signal, and perform weighting processing on the sound signal to be processed according to the power spectral density to obtain a spectrum estimate of the target sound signal. Specifically, after determining the power spectral density of the interfering sound signal, a posteriori and a priori signal-to-noise ratios are determined, and the weighting coefficients are calculated according to the signal-to-noise ratios, and the to-be-processed sound signal is weighted to obtain a spectrum estimate of the target sound information.

Step S13: Determine the masking threshold according to the spectrum estimation.

Step S14: When it is determined that the spectral component of the interfering sound signal in the sound signal to be processed is greater than the masking threshold, the sound signal to be processed is filtered.

And, in this embodiment of the present invention, for the calculation of the masking threshold, specifically:

According to the spectrum estimation, determine the power spectral density B(k) of the critical band of the sound signal to be processed and the extended critical band spectrum C(k):

C(k)=B(k)*SF(k),

Among them, SF(k)=15.81+7.5*k+0.474-17.5*sqrt(1+(k+0.474)2), bh, bl are the upper and lower limit frequencies of each critical band respectively;

According to the extended critical band spectrum C(k) and the offset function O(k), determine the preliminary masking threshold T(k):

T(k)=10 ^{lg(C(k))-(O(k)/10)} ,

Among them, the offset function O(k)=belta*(14.5+k)+(1-belta)*5.5; belta is the pitch coefficient;

From the preliminary masking threshold T(k) and the absolute hearing threshold T _abs (k), determine the masking threshold R _TT (Ω):

R _TT (Ω)=min(T(k),T _abs (k)),

Wherein, T _abs (k)=3.64f ^-0.8 -6.5exp(f-3.3) ² +10 ^-3 f ⁴ .

In the embodiment of the present invention, the masking threshold is further determined through the calculated power spectral density PSD of the interfering sound signal, and this process reduces the computational complexity of the algorithm. Moreover, the order requirement of the pre-echo canceller is reduced, thereby accelerating the convergence speed of the pre-echo canceller. And, it can improve its robustness in strong background noise and near-end speech environment.

As shown in FIG. 2, an embodiment of the present invention provides a method for processing a sound signal, including:

Step S21: Receive an initial sound signal. The initial sound signal can be picked up by a microphone and other audio equipment.

Step S22: Perform echo cancellation on the initial sound signal by an echo canceller to obtain the sound signal to be processed.

Step S23: Determine the power spectral density of the interfering sound signal, and perform weighting processing on the sound signal to be processed according to the power spectral density to obtain a spectrum estimate of the target sound signal.

Step S24: Determine the masking threshold according to the spectrum estimation.

Step S25: When it is determined that the spectral component of the interfering sound signal in the sound signal to be processed is greater than the masking threshold, the sound signal to be processed is filtered.

In the embodiment of the present invention, after the initial signal is received, initial echo cancellation is performed on it, which can improve the processing accuracy of the sound signal.

If the sound signal to be processed includes noise signals and echo signals, weighting processing is performed on the sound signal to be processed according to the power spectral density to obtain the spectral estimation of the target sound signal:

Convert the sound signal to be processed into a frequency domain signal E(Ω);

Determine the posterior signal-to-noise ratio PostSNR(Ω) according to the following formula:

PostSNR(Ω)=|E(Ω)| ² /(R _bb (Ω)+R _nn (Ω)),

Among them, R _bb (Ω) is the power spectral density of the echo signal, and R _nn (Ω) is the power spectral density of the noise signal;

The priori SNR(Ω) is derived from the following formula:

PrioriSNR(Ω _i )=(1-alpha)*P(PostSNR(Ω _i )-1)+alpha*|S'(Ω _i-1 )| ² /R _bb (Ω);

Among them, alpha is the smoothing factor, P(x)=(|x|+x)/2, S'(Ω _i-1 ) is the spectral estimation of the sound signal of the previous frame;

Further calculate the weighting coefficient H _LSA (Ω), and obtain the spectral estimation S'(Ω) of the target sound signal:

S'(Ω)=E(Ω)*H _LSA (Ω),

Wherein, theta=PostSNR(Ω)*PrioriSNR(Ω)/(PrioriSNR(Ω)+1).

When it is determined that the spectral component of the interfering sound signal in the sound signal to be processed is greater than the masking threshold, the steps of filtering the sound signal to be processed include:

According to the power spectral density of the echo signal and the power spectral density of the noise signal, the weighting coefficient H(Ω) of the filtering process is determined:

H(Ω)=min(1, sqrt(R _TT (Ω)/(R _bb (Ω)+R _nn (Ω))) +(zeta_b*R _bb (Ω)+zeta_n*R _nn (Ω))/ (R _bb (Ω)+R _nn (Ω))),

Among them, R _bb (Ω) is the power spectral density of the echo signal, R _nn (Ω) is the power spectral density of the noise signal, zeta_b is the echo attenuation coefficient, and zeta_n is the noise attenuation coefficient.

In the embodiment of the present invention, the original background noise characteristics are retained, the residual echo auditory test is more like noise, and the speech distortion is reduced, so that the voice sounds more natural. Moreover, the order requirement of the pre-echo canceller is reduced, thereby accelerating the convergence speed of the pre-echo canceler and reducing the computational complexity of the echo canceler algorithm. And, it can improve its robustness in strong background noise and near-end speech environment.

As shown in Fig. 3, in the embodiment of the present invention, the method for processing a voice signal of the present invention realizes that the voice signal transmitted from the far-end microphone in the system is shown by the speaker, and an initial echo signal d(k) is formed. The near-end microphone picks up the speech signal y(k), which includes the pure speech signal s(k), the target sound signal, the noise signal n(k), and the original echo signal d(k) fed back by the speaker through the LRM. First, the echo canceler C performs echo cancellation on the speech signal y(k) picked up by the near-end microphone, and the filter H further performs filtering processing.

As shown in FIG. 4, an embodiment of the present invention provides a method for processing a voice signal, including:

The near-end microphone picks up the speech signal y(k), which includes the pure speech signal s(k), the noise signal n(k), and the original echo signal d(k) fed back by the speaker via the LRM. In this embodiment of the present invention, the pure voice signal is target information.

The echo canceller performs echo cancellation on the voice signal y(k) picked up by the near-end microphone, and obtains the voice signal e(k) after echo cancellation. The interference sound signals included in the echo-cancelled speech signal e(k) are noise signals and residual echo signals.

Noise PSD Rnn (Ω) and residual echo PSD R _bb (Ω) are estimated by statistical or autocorrelation methods.

The post-filter performs weighting processing on the near-end microphone signal after echo cancellation, and obtains the preliminary estimation S'(Ω) of the spectrum of the pure voice signal. The specific process includes:

a) Calculate the posterior signal-to-noise ratio:

PostSNR(Ω)＝|E(Ω)| ² /(R _bb (Ω)+R _nn (Ω))

b) Derive the prior signal-to-noise ratio according to the decision-guided method:

PrioriSNR(Ω _i )=(1-alpha)*P(PostSNR(Ω _i )-1)+alpha*|S'(Ω _i-1 )| ² /R _bb (Ω)

where alpha is a smoothing factor, P(x)=(|x|+x)/2, S'(Ω _i-1 ) is a preliminary estimation of the speech signal of the previous frame.

c) Define theta=PostSNR(Ω)*PrioriSNR(Ω)/(PrioriSNR(Ω)+1), and then calculate the weighting coefficient:

d) Preliminary estimation of speech signal obtained by weighting S'(Ω)=E(Ω)*H _LSA (Ω)

Then, the masking threshold R _TT (Ω) is estimated according to the preliminary estimation S' (Ω) of the speech signal spectrum. The specific process includes:

a) Perform critical band analysis on the signal. According to the position theory, the human ear is regarded as a discrete band-pass filter bank, and a critical band is called a Bark, then

Power spectral density of each critical band

Among them, bh and bl are the upper and lower limit frequencies of each critical frequency band, respectively, and k is related to the sampling rate.

b) Calculate the spread function SF(k):

SF(k)=15.81+7.5*k+0.474-17.5*sqrt(1+(k+0.474)2)

Due to the interaction between critical bands, the extended extended critical band spectrum can be expressed as C(k)=B(k)*SF(k).

c) Calculate the masking threshold R _TT (Ω) for masking noise and residual echo.

There are two masking thresholds, namely: the threshold for pure tone masking noise and residual echo, which is C(k)-(14.5+k)db, and the threshold for noise and residual echo masking pure tone, which is C(k)-5.5db .

Therefore, to determine whether the signal resembles a pure tone or noise and residual echo, it is necessary to define the spectral flatness measure SFM:

SFM=10*lg(G/A)

Among them, G and A are the geometric mean and arithmetic mean of the signal power spectral density, respectively.

And, define the pitch coefficient belta=min(SFM/SFM _max ,1)

The offset function O(k) of the masking energy of each frequency band is calculated by belta:

O(k)=belta*(14.5+k)+(1-belta)*5.5

Then the masking threshold size is: T(k)=10 ^{lg(C(k))-(O(k)/10)}

Return the computed spread function threshold to the Bark domain

Compared with the absolute hearing threshold of the human ear, if the calculated masking threshold is lower than the absolute hearing threshold of the human ear, the value of the absolute hearing threshold is taken, where the absolute hearing threshold Tabs(k) is defined as:

Tabs(k)=3.64f ^-0.8 -6.5exp(f-3.3) ² +10 ^-3 f ⁴

Therefore, the final masking threshold is R _TT (Ω)=min(T(k),T _abs (k)).

Further, psychoacoustic weighting filtering is performed on the frequency domain microphone signal E(Ω) after echo cancellation. FFT (Fast Fourier Transform) can be used to convert the digital signal in the time domain into a frequency domain signal, and judge whether the noise spectral component in the frequency domain microphone signal E(Ω) after echo cancellation is less than the masking threshold, if so, keep it and not process it; if Otherwise, the corresponding noise spectral components are attenuated according to conventional MMSE-LSA.

Among them, the specific derivation process of the psychoacoustic weighting filter coefficients is as follows:

The design goal of psychoacoustic adaptive weighted filtering is to minimize the distortion of the near-end speech signal when the sum of residual echo distortion and noise distortion is equal to the masking threshold, so the optimal psychoacoustic weighting filter coefficient H(Ω) satisfies:

[zeta_b–H(Ω)]2R _bb (Ω)+[zeta_n–H(Ω)]2R _nn (Ω)=R _TT (Ω)

Among them, zeta_b is the residual echo attenuation coefficient, usually 20lg(zeta_b)=-35;

zeta_n is the noise attenuation coefficient, usually 20lg(zeta_n)=-15.

Since 0<=H(Ω)<=1, solving the above quadratic equation H(Ω) takes a positive value:

H(Ω)=min(1,[zeta_b* _Rbb (Ω)+zeta_n* _Rnn (Ω)+

sqrt([R _bb (Ω)+R _nn (Ω)]*R _TT (Ω)-[zeta_b-zeta_n] ² *R _bb (Ω)*R _bb (Ω))]/(R _bb (Ω)+ R _nn (Ω)))

Since zeta_b and zeta_n are much smaller than 1 and R _TT (Ω) is usually not too small relative to R _bb (Ω) and R _bb (Ω), the above formula can be simplified to:

H(Ω)=min(1, sqrt(R _TT (Ω)/(R _bb (Ω)+R _nn (Ω))) +(zeta_b*R _bb (Ω)+zeta_n*R _nn (Ω))/ (R _bb (Ω)+R _nn (Ω)))

In the embodiment of the present invention, since the psychoacoustic post-filter can also reduce the order requirement of the pre-echo cancellation adaptive filter, the convergence speed of the echo canceller can be accelerated, the computational complexity of the algorithm can be reduced, and the performance of the echo canceller can be improved. Robustness in strong background noise and near-end speech environments.

And, the residual echo cancellation is integrated in the post psychoacoustic weighting filter, and the residual echo is used to adaptively update the filter weighting coefficient to further eliminate the acoustic echo. In addition, the noise spectrum and residual echo components below the masking threshold are inaudible due to the masking effect of the human ear, so this part of the noise spectrum and residual echo components do not need to be attenuated. The noise spectrum masked by the speech signal and the residual echo components are attenuated, so that the original background noise characteristics are well preserved. The residual echo auditory test is more like noise, the speech distortion is reduced, and the sound is more natural.

It should be noted that, for the sake of simple description, the foregoing method embodiments are all expressed as a series of actions combined, but those skilled in the art should know that the present invention is not limited by the described sequence of actions. As in accordance with the present invention, certain steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention. In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For the part that is not described in detail in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

As shown in Figure 5, an embodiment of the present invention also provides a system 500 for processing sound signals, including:

The signal acquisition module 510 is configured to acquire a sound signal to be processed, and the sound signal to be processed includes a target sound signal and an interference sound signal.

The spectrum estimation determining module 520 is configured to determine the power spectral density of the interfering sound signal, and perform weighting processing on the sound signal to be processed according to the power spectral density, so as to obtain a spectrum estimation of the target sound signal.

The masking threshold determination module 530 is configured to determine the masking threshold according to the spectrum estimation.

The filtering processing module 540 is configured to perform filtering processing on the sound signal to be processed when it is determined that the spectral component of the interfering sound signal in the sound signal to be processed is greater than the masking threshold.

Further, the interfering sound signals include noise signals and echo signals.

The spectrum estimation and determination module is further configured to convert the to-be-processed sound signal into a frequency domain signal E(Ω); and to determine the posterior signal-to-noise ratio PostSNR(Ω) according to the following formula:

PostSNR(Ω)=|E(Ω)| ² /(R _bb (Ω)+R _nn (Ω)),

Among them, R _bb (Ω) is the power spectral density of the echo signal, and R _nn (Ω) is the power spectral density of the noise signal;

The priori SNR(Ω) is derived from the following formula:

PrioriSNR(Ω _i )=(1-alpha)*P(PostSNR(Ω _i )-1)+alpha*|S'(Ω _i-1 )| ² /R _bb (Ω);

Among them, alpha is the smoothing factor, P(x)=(|x|+x)/2, S'(Ω _i-1 ) is the spectral estimation of the sound signal of the previous frame;

Further calculate the weighting coefficient H _LSA (Ω), and obtain the spectral estimation S'(Ω) of the target sound signal:

S'(Ω)=E(Ω)*H _LSA (Ω),

Wherein, theta=PostSNR(Ω)*PrioriSNR(Ω)/(PrioriSNR(Ω)+1).

The masking threshold determination module is also used to, according to the spectrum estimation, determine the power spectral density B(k) and the extended critical band spectrum C(k) of the critical frequency band of the sound signal to be processed:

C(k)=B(k)*SF(k),

Among them, SF(k)=15.81+7.5*k+0.474-17.5*sqrt(1+(k+0.474)2), bh, bl are the upper and lower limit frequencies of each critical band respectively;

According to the extended critical band spectrum C(k) and the offset function O(k), determine the preliminary masking threshold T(k):

T(k)=10 ^{lg(C(k))-(O(k)/10)} ,

Among them, the offset function O(k)=belta*(14.5+k)+(1-belta)*5.5; belta is the pitch coefficient;

From the preliminary masking threshold T(k) and the absolute hearing threshold T _abs (k), determine the masking threshold R _TT (Ω):

R _TT (Ω)=min(T(k),T _abs (k)),

Wherein, T _abs (k)=3.64f ^-0.8 -6.5exp(f-3.3) ² +10 ^-3 f ⁴ .

The filtering processing module is also used to determine the weighting coefficient H (Ω) of filtering processing according to the power spectral density of the echo signal and the power spectral density of the noise signal:

H(Ω)=min(1, sqrt(R _TT (Ω)/(R _bb (Ω)+R _nn (Ω))) +(zeta_b*R _bb (Ω)+zeta_n*R _nn (Ω))/ (R _bb (Ω)+R _nn (Ω))),

Among them, R _bb (Ω) is the power spectral density of the echo signal, R _nn (Ω) is the power spectral density of the noise signal, zeta_b is the echo attenuation coefficient, and zeta_n is the noise attenuation coefficient.

The signal acquisition module is also used for receiving the initial sound signal; performing echo cancellation on the initial sound signal to obtain the sound signal to be processed.

In the embodiment of the present invention, the masking threshold is further determined through the calculated power spectral density PSD of the interfering sound signal, and this process reduces the computational complexity of the algorithm. Moreover, the order requirement of the pre-echo canceller is reduced, thereby accelerating the convergence speed of the pre-echo canceller. And, it can improve its robustness in strong background noise and near-end speech environment.

In some embodiments, embodiments of the present invention provide a non-volatile computer-readable storage medium, where one or more programs including execution instructions are stored in the storage medium, and the execution instructions can be read by an electronic device (including But it is not limited to a computer, a server, or a network device, etc.) to read and execute it, so as to execute any of the above-mentioned methods for processing a sound signal of the present invention.

In some embodiments, embodiments of the present invention further provide a computer program product, the computer program product including a computer program stored on a non-volatile computer-readable storage medium, the computer program including program instructions, when all When the program instructions are executed by a computer, the computer is made to execute any one of the above-mentioned methods for processing sound signals.

In some embodiments, embodiments of the present invention further provide an electronic device, which includes: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores data that can be accessed by the at least one processor. Instructions executed by one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a method of processing a sound signal.

In some embodiments, embodiments of the present invention further provide a storage medium on which a computer program is stored, characterized in that, when the program is executed by a processor, a method for processing a sound signal.

The system for processing a sound signal according to the above embodiment of the present invention can be used to execute the method for processing a sound signal according to the embodiment of the present invention, and correspondingly achieve the technical effects achieved by the method for processing a sound signal according to the above embodiment of the present invention, which is not repeated here. Repeat. In this embodiment of the present invention, relevant functional modules may be implemented by a hardware processor (hardware processor).

Fig. 6 is the hardware structure schematic diagram of the electronic device that performs the method for processing sound signal provided by another embodiment of the present application, as shown in Fig. 6, this device comprises:

One or more processors 610 and memory 620, one processor 610 is taken as an example in FIG. 6 .

The apparatus for performing the method of processing a sound signal may further include: an input device 630 and an output device 640.

The processor 610, the memory 620, the input device 630 and the output device 640 may be connected by a bus or in other ways, and the connection by a bus is taken as an example in FIG. 6 .

As a non-volatile computer-readable storage medium, the memory 620 can be used to store non-volatile software programs, non-volatile computer-executable programs and modules, such as those corresponding to the methods for processing sound signals in the embodiments of the present application. Program instructions/modules. The processor 610 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 620, that is, implementing the method for processing sound signals in the above method embodiments.

The memory 620 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the apparatus for processing sound signals, and the like. Additionally, memory 620 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, the memory 620 may optionally include memory located remotely from the processor 610, and these remote memories may be connected via a network to the apparatus for processing sound signals. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The input device 630 may receive input numerical or character information, and generate signals related to user settings and function control of the device for processing sound signals. The output device 640 may include a display device such as a display screen.

The one or more modules are stored in the memory 620, and when executed by the one or more processors 610, perform the method for processing a sound signal in any of the above method embodiments.

The above product can execute the method provided by the embodiments of the present application, and has corresponding functional modules and beneficial effects for executing the method. For technical details not described in detail in this embodiment, reference may be made to the methods provided in the embodiments of this application.

The electronic devices of the embodiments of the present application exist in various forms, including but not limited to:

(1) Mobile communication equipment: This type of equipment is characterized by having mobile communication functions, and its main goal is to provide voice and data communication. Such terminals include: smart phones (eg iPhone), multimedia phones, feature phones, and low-end phones.

(2) Ultra-mobile personal computer equipment: This type of equipment belongs to the category of personal computers, has computing and processing functions, and generally has the characteristics of mobile Internet access. Such terminals include: PDAs, MIDs, and UMPC devices, such as iPads.

(3) Portable entertainment equipment: This type of equipment can display and play multimedia content. Such devices include: audio and video players (eg iPod), handheld game consoles, e-books, as well as smart toys and portable car navigation devices.

(4) Server: a device that provides computing services. The composition of the server includes a processor, a hard disk, a memory, a system bus, etc. The server is similar to a general computer architecture, but due to the need to provide highly reliable services, it has a great impact on processing capacity, stability, etc. , reliability, security, scalability, manageability and other aspects of high requirements.

(5) Other electronic devices with data interaction function.

The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

From the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, and certainly can also be implemented by hardware. Based on this understanding, the above-mentioned technical solutions can be embodied in the form of software products in essence, or the parts that make contributions to related technologies, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic disks , optical disc, etc., including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the methods described in various embodiments or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still be The technical solutions described in the foregoing embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions in the embodiments of the present application.

Claims (11)

1. A method of processing a sound signal, comprising:

acquiring a sound signal to be processed, wherein the sound signal to be processed comprises a target sound signal and an interference sound signal, and the interference sound signal comprises a noise signal and an echo signal;

determining a power spectral density of the interfering sound signal;

converting the sound signal to be processed into a frequency domain signal E (omega);

the posterior signal-to-noise ratio PostSNR (Ω) is determined according to the following formula:

PostSNR(Ω)＝|E(Ω)|²/(R_bb(Ω)+R_nn(Ω))，

wherein R is_bb(omega) is the power spectral density, R, of the echo signal_nn(Ω) is the power spectral density of the noise signal;

the a priori signal-to-noise ratio PrioriSNR (Ω) is derived according to the following equation:

PrioriSNR(Ω_i)＝(1-alpha)*P(PostSNR(Ω_i)-1)+alpha*|S’(Ω_i-1)|²/R_bb(Ω)；

where alpha is a smoothing factor, p (x) ═ (| x | + x)/2, S' (Ω)_i-1) Estimating the frequency spectrum of the sound signal of the previous frame;

further calculating a weighting factor H_LSA(Ω), and obtaining a spectral estimate S' (Ω) of the target sound signal:

S’(Ω)＝E(Ω)*H_LSA(Ω)，

wherein theta is PostSNR (Ω) × PrioriSNR (Ω)/(PrioriSNR (Ω) + 1);

determining a masking threshold from the spectral estimate;

and under the condition that the frequency spectrum component of the interference sound signal in the sound signal to be processed is determined to be larger than the masking threshold, carrying out filtering processing on the sound signal to be processed.

2. The method according to claim 1, wherein the step of performing filtering processing on the sound signal to be processed when it is determined that the spectral component of the interfering sound signal in the sound signal to be processed is greater than the masking threshold comprises:

determining a weighting coefficient H (omega) of the filtering process according to the power spectral density of the echo signal and the power spectral density of the noise signal:

H(Ω)＝min(1,sqrt(R_TT(Ω)/(R_bb(Ω)+R_nn(Ω)))+(zeta_b*R_bb(Ω)+zeta_n*R_nn(Ω))/(R_bb(Ω)+R_nn(Ω)))，

wherein R is_bb(omega) is the power spectral density, R, of the echo signal_nn(Ω) is the power spectral density of the noise signal, zeta _ b is the echo attenuation coefficient, and zeta _ n is the noise attenuation coefficient.

3. The method of claim 1, wherein the step of determining a masking threshold based on the spectral estimate comprises:

determining, from the spectral estimation, a power spectral density b (k) and an extended critical band spectrum c (k) of a critical band of the sound signal to be processed:

C(k)＝B(k)*SF(k)，

wherein sf (k) ═ 15.81+7.5 × k +0.474-17.5 × sqrt (1+ (k +0.474)2), bh, bl are the upper and lower limit frequencies of each critical band, respectively;

determining a preliminary masking threshold t (k) according to the spread critical band spectrum c (k) and the offset function o (k):

T(k)＝10^{lg(C(k))-(O(k)/10)}，

wherein the offset function o (k) ═ belta (14.5+ k) + (1-belta) × 5.5; belta is the pitch coefficient;

according to a preliminary masking threshold T (k) and an absolute hearing threshold T_abs(k) Determining a masking threshold R_TT(Ω)：

R_TT(Ω)＝min(T(k),T_abs(k))，

Wherein, T_abs(k)＝3.64f^-0.8-6.5exp(f-3.3)²+10^-3f⁴。

4. The method of claim 1, wherein the step of obtaining the sound signal to be processed comprises:

receiving an initial sound signal;

and carrying out echo cancellation on the initial sound signal to obtain the sound signal to be processed.

5. The method according to claim 1, characterized in that the sound signal to be processed is a speech signal.

6. A system for processing a sound signal, comprising:

the device comprises a signal acquisition module, a processing module and a processing module, wherein the signal acquisition module is used for acquiring a sound signal to be processed, the sound signal to be processed comprises a target sound signal and an interference sound signal, and the interference sound signal comprises a noise signal and an echo signal;

the frequency spectrum estimation determining module is used for determining the power spectral density of the interference sound signal and carrying out weighting processing on the sound signal to be processed according to the power spectral density to obtain the frequency spectrum estimation of the target sound signal;

a masking threshold determination module for determining a masking threshold from the spectral estimate;

the filtering processing module is used for performing filtering processing on the sound signal to be processed under the condition that the frequency spectrum component of the interference sound signal in the sound signal to be processed is determined to be larger than the masking threshold;

the frequency spectrum estimation determining module is further used for converting the sound signal to be processed into a frequency domain signal E (omega); and determining the posterior signal-to-noise ratio PostSNR (Ω) according to the following formula:

PostSNR(Ω)＝|E(Ω)|²/(R_bb(Ω)+R_nn(Ω))，

wherein R is_bb(omega) is the power spectral density, R, of the echo signal_nn(Ω) is the power spectral density of the noise signal;

the a priori signal-to-noise ratio PrioriSNR (Ω) is derived according to the following equation:

PrioriSNR(Ω_i)＝(1-alpha)*P(PostSNR(Ω_i)-1)+alpha*|S’(Ω_i-1)|²/R_bb(Ω)；

where alpha is a smoothing factor, p (x) ═ (| x | + x)/2, S' (Ω)_i-1) Estimating the frequency spectrum of the sound signal of the previous frame;

further calculating a weighting factor H_LSA(Ω), and obtaining a spectral estimate S' (Ω) of the target sound signal:

S’(Ω)＝E(Ω)*H_LSA(Ω)，

where theta is PostSNR (Ω) × PrioriSNR (Ω)/(PrioriSNR (Ω) + 1).

7. The system according to claim 6, wherein the masking threshold determination module is further configured to determine, according to the spectral estimation, the power spectral density B (k) and the spectrum C (k) of the critical band of the sound signal to be processed:

C(k)＝B(k)*SF(k)，

wherein sf (k) ═ 15.81+7.5 ═ k +0.474 to 17.5 · sqrt (1+ (k +0.474)²) Bh and bl are the upper and lower limit frequencies of each critical frequency band respectively;

determining a preliminary masking threshold t (k) according to the spread critical band spectrum c (k) and the offset function o (k):

T(k)＝10^{lg(C(k))-(O(k)/10)}，

wherein the offset function o (k) ═ belta (14.5+ k) + (1-belta) × 5.5; belta is the pitch coefficient;

according to a preliminary masking threshold T (k) and an absolute hearing threshold T_abs(k) Determining a masking threshold R_TT(Ω)：

R_TT(Ω)＝min(T(k),T_abs(k))，

Wherein, T_abs(k)＝3.64f^-0.8-6.5exp(f-3.3)²+10^-3f⁴。

8. The system of claim 6, wherein the filtering module is further configured to determine a weighting coefficient H (Ω) of the filtering process according to the power spectral density of the echo signal and the power spectral density of the noise signal:

H(Ω)＝min(1,sqrt(R_TT(Ω)/(R_bb(Ω)+R_nn(Ω)))+(zeta_b*R_bb(Ω)+zeta_n*R_nn(Ω))/(R_bb(Ω)+R_nn(Ω)))，

wherein R is_bb(omega) is the power spectral density, R, of the echo signal_nn(Ω) is the power spectral density of the noise signal, zeta _ b is the echo attenuation coefficient, and zeta _ n is the noise attenuation coefficient.

9. The system of claim 6, wherein the signal acquisition module is further configured to receive an initial sound signal; and carrying out echo cancellation on the initial sound signal to obtain the sound signal to be processed.

10. An electronic device, comprising: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the steps of the method of any one of claims 1-5.

11. A storage medium on which a computer program is stored which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 5.

Images