Static noise reduction method, device, computer equipment and storage medium

By converting the time domain signal of the input voice into a frequency domain signal, confirming the energy and noise energy, and adjusting the target gain algorithm according to the preset parameters, the problem of difficulty in adjusting the gain algorithm and narrow range in the existing technology is solved, and flexible Noise reduction gain adjustment and strength satisfaction.

CN114005456BActive Publication Date: 2024-11-08SHENZHEN WONDERSHARE SOFTWARE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111301460.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-11-04
Publication Date
2024-11-08
Estimated Expiration
2041-11-04

AI Technical Summary

Technical Problem

The existing static noise reduction method needs to be reselected when adjusting the gain algorithm, and the gain range is narrow, making it difficult to flexibly adjust the noise reduction intensity.

Method used

By obtaining the time domain signal of the input voice and converting it into a frequency domain signal, confirming the input energy and static noise energy, obtaining the preset gain algorithm, and adjusting the target gain algorithm according to the preset parameter values ​​to calculate the noise reduction gain, and flexibly adjusting the gain range.

Benefits of technology

It realizes the flexibly adjusting the specific value of the noise reduction gain without changing the existing gain algorithm to meet the noise reduction intensity of different needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114005456B_ABST
    Figure CN114005456B_ABST
Patent Text Reader

Abstract

The present invention discloses a static noise reduction method, device, computer equipment and storage medium, which includes: obtaining an input time domain signal of an input voice, and converting the input time domain signal into an input frequency domain signal; confirming input energy and static noise energy according to the input frequency domain signal, and obtaining a preset gain algorithm, wherein the preset gain algorithm includes a plurality of parameters for changing the gain range; confirming a preset parameter value, and assigning the preset parameter value to a parameter in the preset gain algorithm to obtain a target gain algorithm; calculating a noise reduction gain according to the target gain algorithm, the input energy and the static noise energy; and processing the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal. The present invention can not only improve the noise reduction effect, but also flexibly adjust the gain range.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of audio noise reduction, and in particular to a static noise reduction method, device, computer equipment and storage medium. Background Art

[0002] Static noise is a common type of noise, which often occurs in stable background noises such as electronic noise, engine noise, and cicada sounds. At present, the method of removing static noise is usually to remove the noise energy. In the process of removing the noise energy, it is necessary to balance the noise reduction intensity and the degree of speech retention through the gain algorithm. Common gain algorithms include Wiener filter function, spectrum reduction function and other functions. However, after determining the gain algorithm, usually only the determined gain algorithm can be used for noise reduction. When the gain algorithm needs to be replaced, the gain algorithm needs to be reselected, and the adjustable gain range of the existing gain algorithm for noise reduction is narrow. Summary of the invention

[0003] The embodiments of the present invention provide a static noise reduction method, device, computer equipment and storage medium, which can not only remove the static noise in speech but also flexibly adjust the gain range.

[0004] In a first aspect, an embodiment of the present invention provides a method for reducing static noise, the method comprising:

[0005] Acquire an input time domain signal of an input speech, and convert the input time domain signal into an input frequency domain signal;

[0006] Confirming input energy and static noise energy according to the input frequency domain signal, and acquiring a preset gain algorithm, wherein the preset gain algorithm includes a plurality of parameters for changing a gain range;

[0007] Confirming a preset parameter value, and assigning the preset parameter value to a parameter in the preset gain algorithm to obtain a target gain algorithm;

[0008] Calculating a noise reduction gain according to the target gain algorithm, the input energy, and the static noise energy;

[0009] The input frequency domain signal is processed according to the noise reduction gain to obtain an output time domain signal.

[0010] In a second aspect, an embodiment of the present invention further provides a noise reduction device, the device comprising:

[0011] A signal conversion unit, used to obtain an input time domain signal of the input speech, and convert the input time domain signal into an input frequency domain signal;

[0012] A speech energy confirmation unit, used to confirm input energy and static noise energy according to the input frequency domain signal, and obtain a preset gain algorithm, wherein the preset gain algorithm includes a plurality of parameters for changing a gain range;

[0013] A parameter confirmation unit, used for confirming a preset parameter value, and assigning the preset parameter value to a parameter in the preset gain algorithm to obtain a target gain algorithm;

[0014] A noise reduction gain calculation unit, which calculates the noise reduction gain according to the target gain algorithm, the input energy and the static noise energy;

[0015] The output time domain signal confirmation unit is used to process the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal.

[0016] In a third aspect, an embodiment of the present invention further provides a computer device, which includes a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the above method when executing the computer program.

[0017] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program can implement the above method when executed by a processor.

[0018] The embodiment of the present invention provides a static noise reduction method, device, computer equipment and storage medium. The method includes: obtaining an input time domain signal of an input speech, and converting the input time domain signal into an input frequency domain signal; confirming input energy and static noise energy according to the input frequency domain signal, and obtaining a preset gain algorithm, wherein the preset gain algorithm includes a plurality of parameters for changing the gain range; confirming a preset parameter value, and assigning the preset parameter value to a parameter in the preset gain algorithm to obtain a target gain algorithm; calculating a noise reduction gain according to the target gain algorithm, the input energy and the static noise energy; and processing the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal. When obtaining an input time domain signal, the embodiment of the present invention converts the input time domain signal into an input frequency domain signal, and confirms the input energy and static noise energy according to the input frequency domain signal, then obtains a preset gain algorithm, and confirms the preset parameter value input in advance by the user, assigns the preset parameter value to the preset gain algorithm to obtain a target gain algorithm, and then substitutes the input energy and static noise energy into the target gain algorithm to calculate the noise reduction gain, that is, different gain ranges have different corresponding parameter values, so that the specific value of the noise reduction gain can be flexibly adjusted, and finally the input frequency domain signal is processed with the noise reduction gain to obtain an output time domain signal to achieve the purpose of noise reduction. Based on the above method, not only noise reduction is achieved, but also the gain range can be flexibly adjusted to adjust the noise reduction intensity to meet different needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0020] Figure 1 is a flow chart of a static noise reduction method provided by an embodiment of the present invention;

[0021] Figure 2 is a schematic block diagram of a noise reduction device provided by an embodiment of the present invention;

[0022] Figure 3 It is a schematic block diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, operations, elements, components and / or groups thereof.

[0025] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, the singular forms of "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in the specification of the present invention and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0026] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a static noise reduction method provided by an embodiment of the present invention. Figure 1 As shown, the method includes steps S110 to S150.

[0027] S110, obtaining an input time domain signal of an input speech, and converting the input time domain signal into an input frequency domain signal.

[0028] In an embodiment of the present invention, time domain and frequency domain are basic properties of signals. Time domain describes the relationship between physical signals and time, and frequency domain is a coordinate system used to describe the characteristics of physical signals in terms of frequency. Therefore, in order to facilitate noise reduction, when the input time domain signal of the input speech is obtained, the input time domain signal is converted into an input frequency domain signal. The conversion methods include but are not limited to short-time Fourier transform, improved discrete cosine transform, discrete transform and other methods.

[0029] In some embodiments, such as the present embodiment, the step S110 may include the following steps: acquiring the input speech and extracting an input time domain signal of the input speech; and converting the input time domain signal into the input frequency domain signal based on short-time Fourier transform.

[0030] In an embodiment of the present invention, the input time domain signal can be converted into an input frequency domain signal through short-time Fourier transform. For example, the input time domain signal is x(n), and the input time domain signal x(n) is converted into an input frequency domain signal X(k, m) through short-time Fourier transform, wherein n is a discrete time index, k is a discrete spectrum index, and m is a short-time Fourier transform time index.

[0031] S120, confirming input energy and static noise energy according to the input frequency domain signal, and acquiring a preset gain algorithm, wherein the preset gain algorithm includes a plurality of parameters for changing a gain range.

[0032] In some embodiments, such as the present embodiment, the step S120 may include the following steps: confirming the input energy and static noise energy of the input frequency domain signal based on a voice activity detection algorithm; and acquiring the preset gain algorithm.

[0033] In another embodiment, the step S120 may include the following steps: acquiring a speech probability estimation algorithm; and confirming the input energy and the static noise energy based on the speech probability estimation algorithm.

[0034] In an embodiment of the present invention, the input energy and static noise energy of the input frequency domain signal can be confirmed by a voice activity detection algorithm, which can be specifically obtained by a voice probability estimation algorithm or a voice activity detection algorithm, wherein the voice probability estimation algorithm is also a noise estimation algorithm, including a recursive average noise algorithm, a minimum tracking algorithm and a histogram noise estimation algorithm, which are used to estimate the energy of noise; the voice activity detection algorithm (VAD) is used to detect the state of speech and determine whether the speech is in a silent state or an active state, that is, to separate speech signals and non-speech signals. Since the input time domain signal has been converted into an input frequency domain signal, the silent state and the active state can be determined by the frequency, and then the noise energy is calculated according to the frequency.

[0035] S130, confirming a preset parameter value, and assigning the preset parameter value to a parameter in the preset gain algorithm to obtain a target gain algorithm.

[0036] In the embodiment of the present invention, input energy refers to speech energy, static noise energy is a part of speech energy, and the purpose of noise reduction is to reduce the proportion of static noise energy in speech energy. Methods for obtaining input energy and static noise energy include but are not limited to speech probability estimation algorithms and voice activity detection algorithms, and input energy and static noise energy can be obtained through the above two algorithms. The preset gain algorithm can be:

[0037]

[0038] in, α, β and γ are the parameters, W is the noise reduction gain, is the input energy, is the static noise energy, when α is larger, the noise reduction is smaller, and when β and γ are larger, the noise reduction is greater. That is, by adjusting the specific values ​​of the above three parameters, different target gain algorithms can be confirmed, thereby adjusting the size of the noise reduction gain, and then adjusting the noise reduction degree.

[0039] For example, when α=β=γ=1, substituting α, β and γ into the preset gain algorithm yields:

[0040]

[0041]

[0042] That is, the target gain algorithm is the Wiener filtering algorithm.

[0043] When α=β=1, γ=1 / μ, substituting α, β and γ into the preset gain algorithm yields:

[0044]

[0045]

[0046] That is, the target gain algorithm is a Wiener filter algorithm that can adjust the degree of speech loss.

[0047] When α=1, β=1 / 2, γ=1, substituting α, β and γ into the preset gain algorithm yields:

[0048]

[0049]

[0050] That is, the target gain algorithm is the energy reduction spectrum algorithm.

[0051] It can be concluded from the above three examples that different values ​​of α, β and γ can obtain different target gain functions, and each target gain function has a different gain range, so that the degree of noise reduction can be flexibly adjusted, and the target gain function can be stored in the database in advance, and the preset parameter value is matched with the target gain function. When the user performs noise reduction, the preset parameter value will be input in advance, so that when the gain function needs to be used, the target gain function can be obtained according to the preset parameter value input by the user.

[0052] S140: Calculate a noise reduction gain according to the target gain algorithm, the input energy, and the static noise energy.

[0053] In some embodiments, such as the present embodiment, the step S140 may include the following step: smoothing the noise reduction gain.

[0054] In the embodiment of the present invention, after determining the parameters of the target gain algorithm, the input energy and the static noise energy can be substituted into the preset gain algorithm to obtain the noise reduction gain, and then the noise reduction gain is smoothed. For example, when the values ​​of α, β and γ are all 1, the target gain algorithm is the Wiener filter algorithm. If the noise reduction gain is W(k,m), it can be obtained by Smooth the noise reduction gain W(k,m) to obtain the smoothed noise reduction gain

[0055] S150: Process the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal.

[0056] In some embodiments, such as the present embodiment, the step S140 may include the following steps: obtaining an output frequency domain signal according to the noise reduction gain and the input frequency domain signal; and converting the output frequency domain signal into the output time domain signal based on an inverse short-time Fourier transform.

[0057] In an embodiment of the present invention, after the noise reduction gain is calculated, the output frequency domain signal is obtained by multiplying the noise reduction gain by the input frequency domain signal, and then the output frequency domain signal is converted into an output time domain signal according to an inverse short-time Fourier transform to complete the noise reduction processing of the input time domain signal.

[0058] Figure 2 is a schematic block diagram of a noise reduction device 100 provided in an embodiment of the present invention. Figure 2 As shown, corresponding to the above static noise reduction method, the present invention also provides a noise reduction device 100. The noise reduction device 100 includes a unit for executing the above static noise reduction method. Figure 2 The noise reduction device 100 includes a signal conversion unit 110, a speech energy confirmation unit 120, a parameter confirmation unit 130, a noise reduction gain calculation unit 140 and an output time domain signal confirmation unit 150.

[0059] Among them, the signal conversion unit 110 is used to obtain the input time domain signal of the input speech, and convert the input time domain signal into an input frequency domain signal; the speech energy confirmation unit 120 is used to confirm the input energy and static noise energy according to the input frequency domain signal, and obtain a preset gain algorithm, wherein the preset gain algorithm includes multiple parameters for changing the gain range; the parameter confirmation unit 130 is used to confirm the preset parameter value, and assign the preset parameter value to the parameter in the preset gain algorithm to obtain a target gain algorithm; the noise reduction gain calculation unit 140 calculates the noise reduction gain according to the target gain algorithm, the input energy and the static noise energy; the output time domain signal confirmation unit 150 is used to process the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal.

[0060] Another embodiment of the present invention further provides a noise reduction device, which adds a smoothing processing unit on the basis of the above embodiment.

[0061] The smoothing processing unit is used to perform smoothing processing on the noise reduction gain.

[0062] In some embodiments, such as this embodiment, the signal conversion unit 110 includes a first acquisition unit and a first transformation unit.

[0063] Among them, the first acquisition unit is used to acquire the input speech and extract the input time domain signal of the input speech; the first conversion unit is used to convert the input time domain signal into the input frequency domain signal based on short-time Fourier transform.

[0064] In another embodiment, the signal conversion unit 110 includes a second acquisition unit and a second transformation unit.

[0065] The second acquisition unit is used to acquire a speech probability estimation algorithm; and the second transformation unit is used to confirm the input energy and the static noise energy based on the speech probability estimation algorithm.

[0066] In some embodiments, such as this embodiment, the speech energy confirmation unit 120 includes a first confirmation unit and a third acquisition unit.

[0067] The first confirmation unit is used to confirm the input energy and static noise energy of the input frequency domain signal based on the voice activity detection algorithm; and the third acquisition unit is used to acquire the preset gain algorithm.

[0068] In some embodiments, such as this embodiment, the output time domain signal confirmation unit 150 includes an output frequency domain signal acquisition unit and a third transformation unit.

[0069] Among them, the output frequency domain signal acquisition unit is used to obtain the output frequency domain signal according to the noise reduction gain and the input frequency domain signal; the third transformation unit is used to convert the output frequency domain signal into the output time domain signal based on inverse short-time Fourier transform.

[0070] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned noise reduction device and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0071] The above-mentioned noise reduction device can be implemented in the form of a computer program. The computer program can be Figure 3 Runs on the computer device shown.

[0072] See also Figure 3 , Figure 3 This is a schematic block diagram of a computer device provided in an embodiment of the present application. It can be a terminal or a server, wherein the terminal can be an electronic device with communication functions such as a smart phone, a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device. The server can be an independent server or a server cluster composed of multiple servers.

[0073] See also Figure 3 The computer device 500 includes a processor 502 , a memory and an interface 507 connected via a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .

[0074] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. When the computer program 5032 is executed, the processor 502 may execute a static noise reduction method.

[0075] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .

[0076] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a static noise reduction method.

[0077] The interface 505 is used to communicate with other devices. Those skilled in the art will appreciate that Figure 3The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0078] The processor 502 is used to run the computer program 5032 stored in the memory to implement the following steps:

[0079] Acquire an input time domain signal of an input speech, and convert the input time domain signal into an input frequency domain signal;

[0080] Confirming input energy and static noise energy according to the input frequency domain signal, and acquiring a preset gain algorithm, wherein the preset gain algorithm includes a plurality of parameters for changing a gain range;

[0081] Confirming a preset parameter value, and assigning the preset parameter value to a parameter in the preset gain algorithm to obtain a target gain algorithm;

[0082] Calculating a noise reduction gain according to the target gain algorithm, the input energy, and the static noise energy;

[0083] The input frequency domain signal is processed according to the noise reduction gain to obtain an output time domain signal.

[0084] In one embodiment, when the processor 502 implements the step of acquiring the input time domain signal of the input speech and converting the input time domain signal into the input frequency domain signal, the processor 502 specifically implements the following steps:

[0085] Acquire the input speech, and extract an input time domain signal of the input speech;

[0086] Based on short-time Fourier transform, the input time domain signal is converted into the input frequency domain signal.

[0087] In one embodiment, when the processor 502 implements the step of acquiring the input time domain signal of the input speech and converting the input time domain signal into the input frequency domain signal, the processor 502 specifically implements the following steps:

[0088] Obtain a speech probability estimation algorithm;

[0089] The input energy and the static noise energy are determined based on the speech probability estimation algorithm.

[0090] In one embodiment, when the processor 502 implements the step of confirming the input energy and the static noise energy according to the input frequency domain signal and acquiring the preset gain algorithm, the processor 502 specifically implements the following steps:

[0091] Determining input energy and static noise energy of the input frequency domain signal based on a voice activity detection algorithm;

[0092] Acquire the preset gain algorithm.

[0093] In one embodiment, after implementing the step of calculating the noise reduction gain according to the target gain algorithm, the input energy and the static noise energy, the processor 502 further includes the following steps:

[0094] The noise reduction gain is smoothed.

[0095] In one embodiment, when the processor 502 implements the step of processing the input frequency domain signal according to the noise reduction gain to obtain the output time domain signal, the processor 502 specifically implements the following steps:

[0096] Obtaining an output frequency domain signal according to the noise reduction gain and the input frequency domain signal;

[0097] Based on inverse short-time Fourier transform, the output frequency domain signal is converted into the output time domain signal.

[0098] It should be understood that in the embodiment of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (FSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0099] It is understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a storage medium, which is a computer-readable storage medium. The computer program is executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0100] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program. When the computer program is executed by a processor, any embodiment of the above-mentioned static noise reduction method is implemented.

[0101] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which are computer-readable storage media that can store program codes.

[0102] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0103] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0104] The steps in the method of the embodiment of the present invention can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0105] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device to execute all or part of the steps of the method described in each embodiment of the present invention.

[0106] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0107] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0108] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A noise reduction method, It is characterized in that The method comprises: Acquire an input time domain signal of an input speech, and convert the input time domain signal into an input frequency domain signal; Confirming input energy and static noise energy according to the input frequency domain signal, and acquiring a preset gain algorithm, wherein the preset gain algorithm includes multiple parameters; Determine the value of each parameter according to the gain range required by the user; Calculating a noise reduction gain according to the preset gain algorithm, the input energy and the static noise energy; Processing the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal; The step of confirming input energy and static noise energy according to the input frequency domain signal and obtaining a preset gain algorithm includes: Determining input energy and static noise energy of the input frequency domain signal based on a voice activity detection algorithm; Obtain the preset gain algorithm, the preset gain algorithm is: ,in, , α, β and γ are the parameters, W is the noise reduction gain, is the input energy, is the static noise energy.

2. The noise reduction method according to claim 1, It is characterized in that The step of acquiring an input time domain signal of an input speech and converting the input time domain signal into an input frequency domain signal comprises: Acquire the input speech, and extract an input time domain signal of the input speech; Based on short-time Fourier transform, the input time domain signal is converted into the input frequency domain signal.

3. The noise reduction method according to claim 1, It is characterized in that After the step of calculating the noise reduction gain according to the preset gain algorithm, the input energy and the static noise energy, the method further includes: The noise reduction gain is smoothed.

4. The noise reduction method according to claim 1, It is characterized in that The step of processing the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal comprises: Obtaining an output frequency domain signal according to the noise reduction gain and the input frequency domain signal; Based on inverse short-time Fourier transform, the output frequency domain signal is converted into the output time domain signal.

5. The noise reduction method according to claim 1, It is characterized in that The step of confirming input energy and static noise energy according to the input frequency domain signal comprises: Obtain a speech probability estimation algorithm; The input energy and the static noise energy are determined based on the speech probability estimation algorithm.

6. A noise reduction device, It is characterized in that The device comprises: A signal conversion unit, used to obtain an input time domain signal of the input speech, and convert the input time domain signal into an input frequency domain signal; A speech energy confirmation unit, used to confirm input energy and static noise energy according to the input frequency domain signal, and obtain a preset gain algorithm, wherein the preset gain algorithm includes multiple parameters; A parameter confirmation unit, used to confirm the value of each parameter according to the gain range required by the user; A noise reduction gain calculation unit, which calculates the noise reduction gain according to the preset gain algorithm, the input energy and the static noise energy; an output time domain signal confirmation unit, configured to process the input frequency domain signal according to the noise reduction gain to obtain an output time domain signal; Wherein, the speech energy confirmation unit comprises: A first confirmation unit, configured to confirm input energy and static noise energy of the input frequency domain signal based on a voice activity detection algorithm; The third acquisition unit is used to acquire the preset gain algorithm, where the preset gain algorithm is: ,in, , α, β and γ are the parameters, W is the noise reduction gain, is the input energy, is the static noise energy.

7. A computer device, It is characterized in that The computer device comprises a memory and a processor connected to the memory; the memory is used to store a computer program; the processor is used to run the computer program stored in the memory to execute the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium, It is characterized in that The storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 can be implemented.