WO2023070899A1

WO2023070899A1 - Self-adaptive noise reduction method, system and device, and storage medium

Info

Publication number: WO2023070899A1
Application number: PCT/CN2021/138987
Authority: WO
Inventors: 刘际滨
Original assignee: 歌尔科技有限公司
Priority date: 2021-10-28
Filing date: 2021-12-17
Publication date: 2023-05-04
Also published as: CN114120951A

Abstract

The present application discloses a self-adaptive noise reduction method, system and device, and a storage medium. The method comprises the following steps: obtaining current noise audio data; inputting the current noise audio data into a pre-trained audio noise reduction model to obtain a noise reduction parameter corresponding to a current noise scene; and on the basis of a noise reduction mode of a self-adaptive noise reduction device, using the noise reduction parameter to generate a phase-inverted sound wave signal to perform self-adaptive noise reduction on the current noise audio data. The present application solves the problem of low noise recognition rate, and improves the noise reduction effect of the self-adaptive noise reduction device.

Description

Adaptive noise reduction method, system, device and storage medium

This application claims priority to a Chinese patent application filed with the China Patent Office on October 28, 2021, with application number 202111267424.0 and titled "Adaptive Noise Reduction Method, System, Device, and Storage Medium," the entire contents of which are incorporated by reference in this application.

technical field

The present application relates to the field of active noise reduction earphones, in particular to an adaptive noise reduction method, system, device and storage medium.

Background technique

Noise-cancelling headphones are divided into active noise-cancelling headphones and passive noise-cancelling headphones according to the different noise-cancelling principles they adopt. Among them, active noise-cancelling headphones are divided into three types: feed-forward noise reduction, feedback noise reduction, and hybrid noise reduction. Active noise-canceling headphones generally receive ambient noise through one or more microphones, and then generate a signal that is in phase opposite to the noise sound wave through an electronic circuit to eliminate external noise entering the ear canal. However, the noise recognition rate of the prior art is low, which seriously affects the noise reduction effect of the earphone.

Contents of the invention

In view of this, embodiments of the present application provide an adaptive noise reduction method, system, device, and storage medium to solve the problem of low noise recognition rate.

An embodiment of the present application provides an adaptive noise reduction method, the method comprising:

Get the current noise audio data;

Input the current noise audio data into the pre-trained audio noise reduction model to obtain the noise reduction parameters corresponding to the current noise scene;

Based on the noise reduction mode of the adaptive noise reduction device, the anti-phase sound wave signal is generated by using the noise reduction parameters to perform adaptive noise reduction on the current noise audio data.

Optionally, before the step of inputting the current noise audio data into the pre-trained audio noise reduction model, the step of obtaining the noise reduction parameters corresponding to the current noise scene includes:

Build an audio noise reduction model, specifically including:

Obtain scene audio data of different noise scenes;

Convert the scene audio data to the corresponding noise frequency domain response through the preset method;

The noise frequency domain response is debugged through cascaded filters to obtain noise reduction parameters corresponding to different noise scenarios.

Optionally, after the noise frequency domain response is debugged in the form of a cascaded filter, after obtaining the noise reduction parameters corresponding to different noise scenarios, it also includes:

Use the noise reduction parameters to eliminate and verify the scene audio data of different noise scenes, and obtain the noise reduction error value;

The noise reduction parameter is optimized by using the noise reduction error value until the number of reverse iterations reaches the preset number, then the optimization of the noise reduction parameter is stopped, the noise reduction optimization parameter is obtained, and the noise reduction optimization parameter is used as the noise reduction parameter.

Optionally, use the noise reduction error value to optimize the noise reduction parameters until the number of reverse iterations reaches the preset number, then stop optimizing the noise reduction parameters, obtain the noise reduction optimization parameters, and use the noise reduction optimization parameters as the noise reduction parameters, including:

Carry out reverse iteration operation on the noise reduction error value until the number of reverse iterations reaches the preset number, then stop optimizing the noise reduction parameters;

Obtain the noise reduction optimization parameters generated after each reverse iteration operation, and obtain the average value of the noise reduction optimization parameters based on the noise reduction optimization parameters and the number of reverse iterations;

The average value of the noise reduction optimization parameters is used as the noise reduction parameter.

The noise reduction optimization parameters generated after the last reverse iteration operation are obtained, and the noise reduction optimization parameters are used as the noise reduction parameters.

Optionally, the scene audio data is converted into a corresponding noise frequency domain response by a preset method, including:

Convert the audio data of each scene into multiple noise frequency domain responses through a preset method;

The average value of multiple noise frequency domain responses is obtained as the noise frequency domain response corresponding to the scene audio data.

Optionally, based on the noise reduction mode of the adaptive noise reduction device, the noise reduction parameters are used to perform adaptive noise reduction on the current noise audio data, including:

If the noise reduction mode of the adaptive noise reduction device is feed-forward noise reduction, then use the feed-forward parameters in the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noise audio data;

If the noise reduction mode of the adaptive noise reduction device is feedback noise reduction, then use the feedback parameters in the noise reduction parameters to generate an inverse sound wave signal to perform adaptive noise reduction on the current noise audio data;

If the noise reduction mode of the adaptive noise reduction device is hybrid noise reduction, the hybrid parameter in the noise reduction parameters is used to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noise audio data.

In order to achieve the above purpose, an adaptive noise reduction system is also provided, and the system includes:

A data acquisition module, configured to acquire current noise audio data;

The noise reduction parameter acquisition module is used to input the current noise audio data into the pre-trained audio noise reduction model to obtain the noise reduction parameters corresponding to the current noise scene;

The adaptive noise reduction module is configured to use the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noise audio data based on the noise reduction mode of the adaptive noise reduction device.

In order to achieve the above object, an adaptive noise reduction device is also provided, the adaptive noise reduction device includes: a memory, a processor, and an adaptive noise reduction method program stored in the memory and operable on the processor, the adaptive noise reduction method When the noise reduction method program is executed by the processor, the steps of any one of the above adaptive noise reduction methods are realized.

In order to achieve the above object, a computer storage medium is also provided, on which an adaptive noise reduction method program is stored. When the adaptive noise reduction method program is executed by a processor, any one of the above adaptive noise reduction methods can be realized. step.

One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: obtain the current noise audio data; input the current noise audio data into the pre-trained audio noise reduction model to obtain the current noise scene corresponding The noise reduction parameters; through the pre-trained audio noise reduction model, improve the noise recognition rate and obtain the noise reduction parameters under the high noise recognition rate for adaptive noise reduction equipment to perform noise reduction;

Based on the noise reduction mode of the adaptive noise reduction device, the anti-phase sound wave signal is generated by using the noise reduction parameters to perform adaptive noise reduction on the current noise audio data; among them, the noise reduction mode based on the adaptive noise reduction device is used in the noise reduction parameters The parameters corresponding to the noise reduction mode accurately generate the anti-phase sound wave signal to complete the accurate adaptive noise reduction for the current noise audio data, thereby improving the noise reduction effect of the adaptive noise reduction device.

Description of drawings

FIG. 1 is a schematic flow chart of a first embodiment of the adaptive noise reduction method of the present application;

FIG. 2 is a schematic flowchart of a second embodiment of the adaptive noise reduction method of the present application;

FIG. 3 is a schematic flowchart of specific implementation steps of step S220 in the second embodiment of the adaptive noise reduction method of the present application;

FIG. 4 is a schematic flowchart of another specific implementation step diagram of step S220 in the second embodiment of the adaptive noise reduction method of the present application;

FIG. 5 is a schematic flowchart of specific implementation steps of step S225' of the adaptive noise reduction method of the present application;

FIG. 6 is a schematic flowchart of another specific implementation step of step S225' of the adaptive noise reduction method of the present application;

FIG. 7 is a schematic flow chart of specific implementation steps of step S222 of the adaptive noise reduction method of the present application;

FIG. 8 is a schematic diagram of the adaptive noise reduction system of the present application;

FIG. 9 is a schematic diagram of an adaptive noise reduction device of the present application.

Detailed ways

It should be understood that the specific embodiments described here are only used to explain the present application, and are not intended to limit the present application.

The main solution of the embodiment of the present application is: obtain the current noise audio data; input the current noise audio data into the pre-trained audio noise reduction model to obtain the noise reduction parameters corresponding to the current noise scene; Noise reduction mode, using noise reduction parameters to generate anti-phase sound wave signals to perform adaptive noise reduction on the current noisy audio data; this application solves the problem of low noise recognition rate and improves the noise reduction effect of adaptive noise reduction equipment.

In order to better understand the above-mentioned technical solution, the above-mentioned technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

Referring to Fig. 1, Fig. 1 is the first embodiment of the adaptive noise reduction method of the present application, the method includes:

Step S110: Acquiring current noise audio data;

Specifically, the audio that can be picked up in the current environment where the adaptive noise reduction device is located is preprocessed to obtain current noise audio data, wherein the noise audio data includes at least information such as different spectral features and specific speech duration.

Step S120: Input the current noise audio data into the pre-trained audio noise reduction model to obtain the noise reduction parameters corresponding to the current noise scene;

Specifically, the pre-trained audio noise reduction model can be a noise reduction model based on a deep neural network (DNN), which can be used for uplink call noise reduction. By designing a special neural network model suitable for the field of speech enhancement, the method based on spectral mapping Perform noise cancellation. The deep neural network realizes the processing output of the original sound through a multi-layer learning network (each network has multiple nodes), and realizes the separation of speech under different input conditions.

Step S130: Based on the noise reduction mode of the adaptive noise reduction device, use the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noisy audio data.

It should be noted that the adaptive noise reduction device in this embodiment can be an earphone, and the earphone can be an in-ear type or a semi-in-ear type. This is not limited; in addition, the adaptive noise reduction device may also be applied to virtual reality (Virtual Reality, or VR), augmented reality (Augmented Reality, or AR) or mediated reality (Mediated Reality, or MR).

Specifically, the adaptive noise reduction device can be used to generate sound wave signals with the same signal strength and opposite directions by using the noise reduction parameters to offset the current noisy audio data, so as to achieve the effect of adaptive noise reduction.

In this embodiment, the current noise audio data is obtained; the current noise audio data is input into the pre-trained audio noise reduction model to obtain the noise reduction parameters corresponding to the current noise scene; through the pre-trained audio noise reduction model, the improved Noise recognition rate and obtain noise reduction parameters under high noise recognition rate for adaptive noise reduction equipment to perform noise reduction;

Referring to FIG. 2, FIG. 2 is the second embodiment of the adaptive noise reduction method of the present application. Before the step of inputting the current noise audio data into the pre-trained audio noise reduction model and obtaining the noise reduction parameters corresponding to the current noise scene, it includes:

Step S210: Acquiring current noise audio data;

Step S220: building an audio noise reduction model;

Specifically, by constructing an audio noise reduction model, it can be a noise reduction model based on a deep neural network (DNN), or a hybrid noise reduction model based on DNN-LSTM, or a combination of multiple neural network models in the prior art. The integrated learning noise reduction model is used to improve the noise recognition effect of the audio noise reduction model, which is not limited here.

It should be noted that in the process of audio noise reduction model mapping learning, in the early stage of algorithm optimization, a large learning rate will accelerate the learning effect, but in the later stage of algorithm optimization, a large learning rate will cause large fluctuations, and there will be a problem of wandering around the optimal value. Therefore, in this embodiment, during the training process of the audio noise reduction model, the learning rate gradually decays as the training progresses.

Specifically, the audio noise reduction model can be deployed on the cloud, or on the adaptive noise reduction device side, that is, the local side, which is not limited here and adjusted according to specific settings.

Step S230: Input the current noise audio data into the pre-trained audio noise reduction model to obtain the noise reduction parameters corresponding to the current noise scene;

Step S240: Based on the noise reduction mode of the adaptive noise reduction device, use the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noisy audio data.

In one of the embodiments, based on the noise reduction mode of the adaptive noise reduction device, noise reduction parameters are used to perform adaptive noise reduction on the current noise audio data, including:

Wherein, in the feedback noise reduction mode and the hybrid noise reduction mode, after the inverse sound wave information is generated, the residual noise signal is obtained, and the residual noise signal is fed back to the audio noise reduction model.

Compared with the first embodiment, the second embodiment includes step S220, other steps have been described in the first embodiment, and will not be repeated here.

In this embodiment, a better noise reduction effect is achieved by constructing an audio noise reduction model to increase the accuracy of noise recognition.

Referring to FIG. 3, FIG. 3 shows the specific implementation steps of step S220 in the second embodiment of the adaptive noise reduction method of the present application, and constructs an audio noise reduction model, including:

Step S221: acquiring scene audio data of different noise scenes;

Specifically, a large amount of scene audio data of different noise scenes can be obtained as a training set to ensure the comprehensiveness of the audio noise reduction model training.

Step S222: Convert the scene audio data into corresponding noise frequency domain responses through a preset method;

Specifically, the scene audio data can be transformed into corresponding noise frequency domain responses through Fourier transform. Wherein, the Fourier transform may be a continuous Fourier transform or a discrete Fourier transform, which is not limited here.

Step S223: Debug the noise reduction parameters through the noise frequency domain response in the form of a cascaded filter to obtain noise reduction parameters corresponding to different noise scenarios.

Specifically, the cascaded filter can be a plurality of filters connected together by cascading, wherein, if more than two identical filters are cascaded, the filtering effect can be enhanced; if more than two different If the filters are cascaded, the filtering frequency domain can be extended. In this embodiment, the above-mentioned cascading manner is not limited, and it can be dynamically adjusted as required.

In this embodiment, the debugging of the corresponding combined noise reduction parameters is performed in the form of cascaded filters, so as to accurately obtain the noise reduction parameters corresponding to different noise scenarios.

Referring to Fig. 4, Fig. 4 is another specific implementation step of step S220 in the second embodiment of the adaptive noise reduction method of the present application. The frequency domain response of the noise is debugged in the form of a cascaded filter to obtain different noise After the step of denoising parameters corresponding to the scene, it also includes:

Step S221': Acquiring scene audio data of different noise scenes;

Step S222': convert the scene audio data into corresponding noise frequency domain responses through a preset method;

Step S223': Debug the noise reduction parameters through the noise frequency domain response in the form of a cascaded filter to obtain noise reduction parameters corresponding to different noise scenarios;

Step S224': use the noise reduction parameters to eliminate and verify the scene audio data of different noise scenes, and obtain the noise reduction error value;

Specifically, the elimination verification may be to use the noise reduction parameters to enable the adaptive noise reduction device to generate sound wave signals with the same intensity but opposite directions to perform noise elimination verification on the scene audio data of different noise scenes.

Step S225': Use the noise reduction error value to optimize the noise reduction parameters until the number of reverse iterations reaches the preset number, then stop optimizing the noise reduction parameters, obtain the noise reduction optimization parameters, and use the noise reduction optimization parameters as the noise reduction parameter.

Specifically, the preset times are not limited here, and are dynamically adjusted according to the specific audio noise reduction model and the quality of the training set.

Specifically, in another embodiment, when the noise reduction error value is within the preset error range, the optimization of the noise reduction parameters can be stopped, the current noise reduction optimization parameters can be obtained, and the noise reduction optimization parameters can be used as the noise reduction parameters. Noise parameter.

Compared with the previous embodiment, this embodiment specifically includes step S224' and step S225', and other steps have been described in the previous embodiment, and will not be repeated here.

In this embodiment, an optimization operation on the noise reduction parameters is added to improve the accuracy of the noise reduction parameters, thereby further ensuring the noise reduction effect of the adaptive noise reduction device.

Referring to Fig. 5, Fig. 5 is the specific implementation steps of step S225' of the self-adaptive noise reduction method of the present application. The noise reduction parameters are optimized by using the noise reduction error value until the number of reverse iterations reaches the preset number of times, then the noise reduction parameters are stopped. Perform optimization to obtain noise reduction optimization parameters, and use the noise reduction optimization parameters as noise reduction parameters, including:

Step S225'-1: Perform reverse iteration operation on the noise reduction error value until the number of reverse iterations reaches the preset number, then stop optimizing the noise reduction parameters;

Specifically, the reverse iteration operation can be implemented based on a reverse iterator; the preset number of times is not limited in this embodiment, and the preset number of times is positively correlated with the optimization effect of the noise reduction parameter, that is, the more the preset times are, the more optimized The higher the effect, but it will cause a waste of computing resources, so the setting of the preset number of times is adjusted according to the noise reduction effect.

Step S225'-2: Obtain the noise reduction optimization parameters generated after each reverse iteration operation, and obtain the average value of the noise reduction optimization parameters based on the noise reduction optimization parameters and the number of reverse iterations;

Specifically, it is also possible to obtain the noise reduction optimization parameters generated after each reverse iteration operation, then remove the maximum and minimum values of the noise reduction optimization parameters, and obtain the average value of the remaining noise reduction optimization parameters to ensure that the noise reduction parameters rationality and correctness.

Step S225'-3: Use the average value of the optimized noise reduction parameters as the noise reduction parameters.

In this embodiment, the average value of the optimized noise reduction parameters is obtained as the noise reduction parameters to make the noise reduction parameters more reasonable, thereby ensuring the noise recognition rate of the audio noise reduction model.

Referring to Fig. 6, Fig. 6 is another specific implementation step of step S225' of the self-adaptive noise reduction method of the present application. The noise reduction parameter is optimized by using the noise reduction error value until the number of reverse iterations reaches the preset number of times, then the noise reduction is stopped. The noise parameters are optimized, the noise reduction optimization parameters are obtained, and the noise reduction optimization parameters are used as the noise reduction parameters, including:

Step S225'-11: Perform reverse iteration operation on the noise reduction error value until the number of reverse iterations reaches the preset number, then stop optimizing the noise reduction parameters;

Step S225'-12: Obtain the noise reduction optimization parameters generated after the last reverse iteration operation, and use the noise reduction optimization parameters as the noise reduction parameters.

Specifically, in this embodiment, the noise reduction optimization parameters generated after the last reverse iteration operation are obtained as the noise reduction parameters to ensure that the noise reduction parameters are optimal values, thereby ensuring the noise recognition rate of the audio noise reduction model, and further improving the automatic Adapt to the noise reduction effect of the noise reduction device.

Referring to Fig. 7, Fig. 7 is the specific implementation steps of step S222 of the adaptive noise reduction method of the present application, which converts the scene audio data into corresponding noise frequency domain responses through a preset method, including:

Step S222-1: converting the audio data of each scene into multiple noise frequency domain responses through a preset method;

Step S222-2: Obtain the average value of multiple noise frequency domain responses as the noise frequency domain response corresponding to the scene audio data.

Specifically, in this embodiment, the average value of multiple noise frequency domain responses is obtained as the noise frequency domain response corresponding to the scene audio data, so that the noise frequency domain response is more reasonable and accurate.

This application also protects an adaptive noise reduction system, said system 02, including:

Data acquisition module 21, for acquiring current noise audio data;

The noise reduction parameter acquisition module 22 is used to input the current noise audio data into the pre-trained audio noise reduction model to obtain the corresponding noise reduction parameters of the current noise scene;

The adaptive noise reduction module 23 is configured to use the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noisy audio data based on the noise reduction mode of the adaptive noise reduction device.

The system shown in FIG. 8 includes a data acquisition module 21, a noise reduction parameter acquisition module 22, and an adaptive noise reduction module 23. The system can execute the methods of the embodiments shown in FIGS. 1 to 7. The parts not described in detail in this embodiment, Reference may be made to relevant descriptions of the embodiments shown in FIGS. 1 to 7 . For the execution process and technical effect of this technical solution, refer to the description in the embodiment shown in FIG. 1 to FIG. 7 , and details are not repeated here.

The present application also protects an adaptive noise reduction device. The adaptive noise reduction device includes: a memory, a processor, and an adaptive noise reduction method program stored on the memory and operable on the processor. When the program of the adaptive noise reduction method is executed by the processor, the steps of any one of the above adaptive noise reduction methods are realized.

An adaptive noise reduction device 10 involved in the present application includes at least one processor 12 and a memory 11 as shown in FIG. 9 .

The processor 12 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor 12 or instructions in the form of software. The above-mentioned processor 12 may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. Various methods, steps, and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, or the like. The software module can be located in a mature storage medium in the field such as random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, register. The storage medium is located in the memory 11, and the processor 12 reads the information in the memory 11, and completes the steps of the above method in combination with its hardware.

It can be understood that the memory 11 in the embodiment of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory can be read-only memory (Read Only Memory, ROM), programmable read-only memory (Programmable ROM, PROM), erasable programmable read-only memory (Erasable PROM, EPROM), electrically erasable In addition to programmable read-only memory (Electrically EPROM, EEPROM) or flash memory. The volatile memory can be Random Access Memory (RAM), which acts as external cache memory. By way of illustration and not limitation, many forms of RAM are available, such as Static Random Access Memory (Static RAM, SRAM), Dynamic Random Access Memory (Dynamic RAM, DRAM), Synchronous Dynamic Random Access Memory (Synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (Double DataRate SDRAM, DDRSDRAM), enhanced synchronous dynamic random access memory (Enhanced SDRAM, ESDRAM), synchronous connection dynamic random access memory (Synch link DRAM, SLDRAM) And Direct Memory Bus Random Access Memory (Direct Rambus RAM, DRRAM). The memory 11 of the system and method described in the embodiments of the present application is intended to include but not limited to these and any other suitable types of memory.

The present application also protects a computer storage medium, on which an adaptive noise reduction method program is stored, and when the adaptive noise reduction method program is executed by a processor, the adaptive noise reduction method described in any one of the above is realized method steps.

Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowcharts and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

It should be noted that, in the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The application can be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means can be embodied by one and the same item of hardware. The use of the words first, second, and third, etc. does not indicate any order. These words can be interpreted as names.

While preferred embodiments of the present application have been described, additional changes and modifications to these embodiments can be made by those skilled in the art once the basic inventive concept is appreciated. Therefore, the appended claims are intended to be construed to cover the preferred embodiment and all changes and modifications which fall within the scope of the application.

Obviously, those skilled in the art can make various changes and modifications to the application without departing from the spirit and scope of the application. In this way, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

An adaptive noise reduction method, characterized in that the method comprises:

Get the current noise audio data;

Inputting the current noise audio data into a pre-trained audio noise reduction model to obtain noise reduction parameters corresponding to the current noise scene;

Based on the noise reduction mode of the adaptive noise reduction device, the noise reduction parameters are used to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noise audio data.
The adaptive noise reduction method according to claim 1, wherein before the step of inputting the current noise audio data into a pre-trained audio noise reduction model and obtaining the noise reduction parameters corresponding to the current noise scene, include:

Construct the audio noise reduction model, specifically including:

Obtain scene audio data of different noise scenes;

Converting the scene audio data into a corresponding noise frequency domain response by a preset method;

The noise frequency domain response is adjusted through cascaded filters to obtain noise reduction parameters corresponding to different noise scenarios.
The adaptive noise reduction method according to claim 2, wherein the noise frequency domain response is adjusted in the form of a cascaded filter to obtain noise reduction parameters corresponding to different noise scenes After the steps, also include:

Using the noise reduction parameters to eliminate and verify the scene audio data of the different noise scenes to obtain a noise reduction error value;

Use the noise reduction error value to optimize the noise reduction parameters until the number of reverse iterations reaches the preset number of times, then stop optimizing the noise reduction parameters, obtain noise reduction optimization parameters, and optimize the noise reduction parameter as the denoising parameter.
The adaptive noise reduction method according to claim 3, wherein the noise reduction parameter is optimized by using the noise reduction error value, until the number of reverse iterations reaches a preset number of times, then stop performing an operation on the The noise reduction parameters are optimized to obtain the noise reduction optimization parameters, and the noise reduction optimization parameters are used as the noise reduction parameters, including:

Perform reverse iteration operations on the noise reduction error value until the number of reverse iterations reaches a preset number of times, then stop optimizing the noise reduction parameters;

Obtaining the noise reduction optimization parameters generated after each reverse iteration operation, and obtaining the average value of the noise reduction optimization parameters based on the noise reduction optimization parameters and the number of reverse iterations;

The average value of the noise reduction optimization parameters is used as the noise reduction parameter.
The adaptive noise reduction method according to claim 3, wherein the noise reduction parameter is optimized by using the noise reduction error value, until the number of reverse iterations reaches a preset number of times, then stop performing an operation on the Optimizing the noise reduction parameters to obtain the noise reduction optimization parameters, and using the noise reduction optimization parameters as the noise reduction parameters, also includes:

Perform reverse iteration operations on the noise reduction error value until the number of reverse iterations reaches a preset number of times, then stop optimizing the noise reduction parameters;

The noise reduction optimization parameters generated after the last reverse iteration operation are obtained, and the noise reduction optimization parameters are used as the noise reduction parameters.
The adaptive noise reduction method according to claim 2, wherein said converting said scene audio data into a corresponding noise frequency domain response by a preset method comprises:

converting the audio data of each scene into a plurality of noise frequency domain responses through a preset method;

Obtain an average value of the plurality of noise frequency domain responses as the noise frequency domain response corresponding to the scene audio data.
The adaptive noise reduction method according to claim 1, wherein the noise reduction mode based on the adaptive noise reduction device uses the noise reduction parameters to perform adaptive noise reduction on the current noise audio data, include:

If the noise reduction mode of the adaptive noise reduction device is feed-forward noise reduction, use the feed-forward parameters in the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noise audio data ;

If the noise reduction mode of the adaptive noise reduction device is feedback noise reduction, using the feedback parameters in the noise reduction parameters to generate an inverse sound wave signal to perform adaptive noise reduction on the current noise audio data;

If the noise reduction mode of the adaptive noise reduction device is a hybrid noise reduction, use the hybrid parameters in the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noise audio data.
An adaptive noise reduction system, characterized in that the system includes:

A data acquisition module, configured to acquire current noise audio data;

The noise reduction parameter acquisition module is used to input the current noise audio data into the pre-trained audio noise reduction model to obtain the noise reduction parameters corresponding to the current noise scene;

The adaptive noise reduction module is configured to use the noise reduction parameters to generate an anti-phase sound wave signal to perform adaptive noise reduction on the current noise audio data based on the noise reduction mode of the adaptive noise reduction device.
An adaptive noise reduction device, characterized in that the adaptive noise reduction device includes: a memory, a processor, and an adaptive noise reduction method program stored in the memory and operable on the processor, the When the adaptive noise reduction method program is executed by the processor, the steps of the adaptive noise reduction method according to any one of claims 1 to 7 are realized.
A computer storage medium, characterized in that an adaptive noise reduction method program is stored on the computer storage medium, and when the adaptive noise reduction method program is executed by a processor, any one of claims 1 to 7 is implemented. The steps of the adaptive noise reduction method.