A method, system, device, and storage medium for correcting microphone audio signals

Abstract

This application relates to the technical field of audio systems, and particularly to a method, system, device, and storage medium for correcting microphone audio signals. This application first obtains an audio correction mode and target parameters to determine overall audio correction parameters; then, in combination with microphone channel information, personalized correction parameters are set for each channel; audio correction processing is performed according to the overall and personalized parameters; subsequently, the quality of the corrected audio is detected. If the detection result differs from the target parameters and exceeds the threshold, audio quality monitoring is triggered, parameter adjustment information is obtained, and the personalized correction parameters are dynamically updated according to this information. Combining the overall and personalized parameters can process audio more comprehensively. At the same time, the personalized parameters are adaptively adjusted, which can continuously optimize the processing effect, improve the adaptive adjustment ability, and maintain an ideal noise reduction ability.

Description

Technical Field

[0001] The present application relates to the technical field of audio systems, and in particular, to a method, system, device, and storage medium for correcting microphone audio signals. Background Art

[0002] With the continuous development of voice interaction technology, microphone arrays are widely used in various voice interaction devices. A microphone array can effectively capture target voice signals while suppressing interference such as environmental noise and echo, and is the key to achieving high-quality voice input. However, the complex actual environment poses severe challenges to the performance of the microphone array, so appropriate correction processing of microphone audio signals is required.

[0003] Existing technologies usually use an adaptive filtering algorithm based on noise spectrum estimation to perform noise suppression processing on microphone audio signals. This type of algorithm first estimates the environmental noise spectrum, then designs an adaptive filter according to the noise spectrum, and filters each microphone channel to suppress noise interference. Existing technologies usually assume that the noise is stationary, but in the actual environment, the noise is often non-stationary. Existing technologies lack the ability of adaptive adjustment and cannot dynamically optimize parameters according to environmental changes, resulting in difficult to guarantee the processing effect in complex environments, and this situation needs to be further improved. Summary of the Invention

[0004] To solve the problem that the existing audio signal correction method lacks the ability of adaptive adjustment, the present application provides a method, system, device, and storage medium for correcting microphone audio signals, and adopts the following technical solutions:

[0005] In a first aspect, the present application provides a method for correcting microphone audio signals, including the following steps:

[0006] Obtain an audio correction mode and target parameters, and obtain overall audio correction parameters according to the audio correction mode and the target parameters;

[0007] Obtain microphone channel information, and set personalized correction parameters for the microphone channels according to the microphone channel information and the overall audio correction parameters;

[0008] Perform audio correction according to the personalized correction parameters and the overall audio correction parameters;

[0009] Perform quality detection on the corrected audio to obtain audio detection parameters; compare the audio detection parameters with the target parameters to obtain a parameter difference value;

[0010] If the parameter difference value exceeds a preset threshold, trigger audio quality monitoring; obtain parameter adjustment information according to the audio quality monitoring result;

[0011] Update the personalized correction parameters according to the parameter adjustment information.

[0012] By adopting the above technical solution, the present application first obtains the audio correction mode and target parameters to determine the overall audio correction parameters; then combines the microphone channel information to set personalized correction parameters for each channel; performs audio correction processing according to the overall and personalized parameters; subsequently, performs quality detection on the corrected audio. If the detection result is different from the target parameters and exceeds the threshold, audio quality monitoring is triggered, parameter adjustment information is obtained, and the personalized correction parameters are dynamically updated according to this information. Combining the overall and personalized parameters can process audio more comprehensively. At the same time, the personalized parameters are adaptively adjusted, which can continuously optimize the processing effect, improve the adaptive adjustment ability, and maintain an ideal noise reduction ability.

[0013] Optionally, the obtaining of the audio correction mode and target parameters, and the obtaining of the overall audio correction parameters according to the audio correction mode and the target parameters specifically include the following steps:

[0014] Extract the target signal-to-noise ratio from the target parameters and use the target signal-to-noise ratio as the main noise suppression target;

[0015] Obtain the noise estimation parameters, adaptive filter parameters, and preprocessing parameters according to the target signal-to-noise ratio and the audio correction mode;

[0016] Associate the target signal-to-noise ratio, noise estimation parameters, adaptive filter parameters, and preprocessing parameters to obtain the overall audio correction parameters.

[0017] By adopting the above technical solution, the present application first extracts the target signal-to-noise ratio from the target parameters as the main noise suppression target; then obtains the noise estimation parameters, adaptive filter parameters, and preprocessing parameters respectively according to the target signal-to-noise ratio and the audio correction mode; finally, associates the target signal-to-noise ratio, noise estimation parameters, filter parameters, and preprocessing parameters to obtain the overall audio correction parameters that meet the requirements.

[0018] Optionally, the preprocessing parameters include pre-emphasis filter parameters and de-pre-emphasis filter parameters, and the noise estimation parameters include noise spectrum estimation parameters and noise tracking factor parameters.

[0019] The obtaining of the noise estimation parameters, adaptive filter parameters, and preprocessing parameters according to the target signal-to-noise ratio and the audio correction mode specifically includes the following steps:

[0020] Obtain the pre-emphasis filter parameters, de-pre-emphasis filter parameters, noise spectrum estimation parameters, noise tracking factor parameters, and adaptive filter parameters according to the target signal-to-noise ratio and the audio correction mode;

[0021] Obtain the noise estimation parameter according to the noise spectrum estimation parameter and the noise tracking factor parameter.

[0022] By adopting the above technical solution, the present application first obtains the pre-emphasis filter parameter, the de-pre-emphasis filter parameter, the noise spectrum estimation parameter, the noise tracking factor parameter, and the adaptive filter parameter according to the target signal-to-noise ratio and the audio correction mode, and then combines the noise spectrum estimation parameter and the noise tracking factor parameter to obtain the final noise estimation parameter; the pre-processing and noise estimation links are subdivided, and the parameters of each sub-link can be finely controlled.

[0023] Optionally, the personalized correction parameter includes a channel noise suppression parameter and a channel feedback suppression parameter. Obtaining the microphone channel information and setting the personalized correction parameter for the microphone channel according to the microphone channel information and the overall audio correction parameter specifically includes the following steps:

[0024] Obtain the microphone channel position information and the channel noise characteristic information;

[0025] Obtain the channel pre-processing parameter for the microphone channel according to the pre-processing parameter, the channel position information, and the channel noise characteristic information;

[0026] Obtain the channel noise suppression parameter and the channel feedback suppression parameter for the microphone channel according to the noise estimation parameter, the adaptive filter parameter, the channel position information, and the channel noise characteristic information.

[0027] By adopting the above technical solution, due to the differences in the position and environment of different microphone channels, there will be certain differences in their noise characteristics and feedback situations. The present application obtains the corresponding channel pre-processing parameter, channel noise suppression parameter, and channel feedback suppression parameter for each channel according to the position information and noise characteristic information of the microphone channel, combined with the pre-processing parameter, noise estimation parameter, and adaptive filter parameter in the overall audio correction parameter, as the personalized correction parameter for the channel, which can more accurately reflect the actual state of each microphone channel and form a good cooperation with the overall strategy, so as to achieve high-quality noise reduction in a complex environment.

[0028] Optionally, the obtaining the parameter adjustment information according to the audio quality monitoring result specifically includes the following steps:

[0029] Obtain the audio quality change information according to the audio quality monitoring result;

[0030] Obtain the quality anomaly trigger information, and trigger the environment detection instruction according to the quality anomaly trigger information;

[0031] Obtain environmental detection result information, and based on the environmental detection result information, in combination with the audio quality change information, obtain the current audio processing status information;

[0032] Based on the current audio processing status information and the overall audio correction parameters, obtain parameter adjustment information.

[0033] By adopting the above technical solution, when the audio quality monitoring finds that the quality index changes, it is necessary to analyze the cause of the change and make corresponding adjustments. This application obtains the current audio processing status information based on the environmental detection result information in combination with the previously obtained audio quality change information. When the audio quality is abnormal, it can actively diagnose the environment, analyze the status, and make targeted parameter adjustments, improving the timeliness and pertinence of the adjustment.

[0034] Optionally, updating the personalized correction parameters according to the parameter adjustment information specifically includes the following steps:

[0035] Based on the parameter adjustment information, obtain personalized parameter adjustment information, where the personalized parameter adjustment information includes the adjusted channel noise suppression parameter and the adjusted channel feedback suppression parameter;

[0036] Compare the adjusted channel noise suppression parameter and the adjusted channel feedback suppression parameter with a preset parameter threshold. If within the threshold range, update the personalized correction parameters according to the personalized parameter adjustment information;

[0037] If exceeding the threshold range, trigger a parameter secondary adjustment instruction, obtain secondary adjustment parameter information according to the parameter secondary adjustment instruction, and update the personalized correction parameters according to the secondary adjustment parameter information.

[0038] By adopting the above technical solution, this application first obtains the adjusted personalized parameters according to the parameter adjustment information, and then compares them with the preset threshold. Only when within the threshold range, the update operation is performed; if the adjusted parameters exceed the expected range, a secondary adjustment procedure will be triggered, and the parameter value will be re-determined according to the secondary adjustment information and then updated, thereby avoiding parameter abnormalities, ensuring that the adjustment is within a controllable range, and improving the stability and processing quality of the overall system.

[0039] In a second aspect, this application provides a microphone audio signal correction system, including:

[0040] An audio correction mode acquisition module, configured to acquire an audio correction mode;

[0041] A target parameter acquisition module, configured to acquire target parameters;

[0042] An overall parameter generation module for obtaining overall audio correction parameters according to the audio correction mode and the target parameters;

[0043] A channel parameter setting module for obtaining microphone channel information and setting personalized correction parameters for each microphone channel according to the microphone channel information and the overall audio correction parameters;

[0044] An audio correction module for performing audio correction according to the personalized correction parameters and the overall audio correction parameters;

[0045] A quality detection module for detecting the quality of the corrected audio and obtaining audio detection parameters;

[0046] A difference acquisition module for comparing the audio detection parameters with the target parameters to obtain a parameter difference value;

[0047] An adaptive adjustment trigger module for triggering audio quality monitoring when the parameter difference value exceeds a preset threshold;

[0048] A parameter adjustment module for obtaining parameter adjustment information according to the adaptive adjustment mechanism;

[0049] A parameter update module for updating the personalized correction parameters according to the parameter adjustment information.

[0050] Optionally, the parameter update module includes:

[0051] A personalized parameter adjustment unit for obtaining personalized parameter adjustment information according to the parameter adjustment information, where the personalized parameter adjustment information includes adjusted channel noise suppression parameters and adjusted channel feedback suppression parameters;

[0052] A parameter threshold comparison unit for comparing the adjusted channel noise suppression parameters and the adjusted channel feedback suppression parameters with preset parameter thresholds;

[0053] A parameter update unit for updating the personalized correction parameters according to the personalized parameter adjustment information when the adjusted parameters are within the threshold range;

[0054] A secondary adjustment trigger unit for triggering a parameter secondary adjustment instruction when the adjusted parameters exceed the threshold range;

[0055] A secondary adjustment parameter acquisition unit for obtaining secondary adjustment parameter information according to the parameter secondary adjustment instruction;

[0056] The parameter update unit is further configured to update the personalized correction parameters according to the secondary adjustment parameter information.

[0057] In a third aspect, the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned microphone audio signal correction method are implemented.

[0058] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned microphone audio signal correction method are implemented.

[0059] In summary, the present application includes at least one of the following beneficial technical effects:

[0060] 1. The present application first obtains an audio correction mode and target parameters to determine overall audio correction parameters; then, in combination with microphone channel information, personalized correction parameters are set for each channel; audio correction processing is performed according to the overall and personalized parameters; subsequently, the quality of the corrected audio is detected. If the detection result differs from the target parameters and exceeds the threshold, audio quality monitoring is triggered, parameter adjustment information is obtained, and the personalized correction parameters are dynamically updated according to this information. Combining the overall and personalized parameters can process audio more comprehensively. At the same time, the personalized parameters are adaptively adjusted, which can continuously optimize the processing effect, improve the adaptive adjustment ability, and maintain an ideal noise reduction ability.

[0061] 2. The present application first obtains pre-emphasis filter parameters, de-pre-emphasis filter parameters, noise spectrum estimation parameters, noise tracking factor parameters, and adaptive filter parameters according to the target signal-to-noise ratio and audio correction mode, and then combines the noise spectrum estimation parameters and noise tracking factor parameters to obtain the final noise estimation parameters; the pre-processing and noise estimation links are subdivided, and the parameters of each sub-link can be finely controlled.

[0062] 3. According to the position information and noise characteristic information of the microphone channels, in combination with the pre-processing parameters, noise estimation parameters, and adaptive filter parameters in the overall audio correction parameters, the corresponding channel pre-processing parameters, channel noise suppression parameters, and channel feedback suppression parameters are obtained for each channel as the personalized correction parameters of the channel, which can more accurately reflect the actual state of each microphone channel and form a good cooperation with the overall strategy, so as to achieve high-quality noise reduction in a complex environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 is a schematic flowchart of a method for correcting a microphone audio signal according to an embodiment of the present application;

[0064] Figure 2 is a schematic flowchart of step S10 in a method for correcting a microphone audio signal according to an embodiment of the present application;

[0065] Figure 3 It is a schematic flowchart of step S22 in a method for correcting a microphone audio signal according to an embodiment of the present application;

[0066] Figure 4 It is a schematic flowchart of step S30 in a method for correcting a microphone audio signal according to an embodiment of the present application;

[0067] Figure 5 It is a schematic flowchart of step S40 in a method for correcting a microphone audio signal according to an embodiment of the present application;

[0068] Figure 6 It is a schematic flowchart of step S60 in a method for correcting a microphone audio signal according to an embodiment of the present application;

[0069] Figure 7 It is a schematic diagram of modules of a microphone audio signal correction system according to an embodiment of the present application;

[0070] Figure 8 It is an internal structure diagram of an electronic device according to an embodiment of the present application. Specific embodiments

[0071] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. As used in the specification and appended claims of the present application, the singular forms "a", "one", "said", "above-mentioned", "this" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to any and all possible combinations including one or more of the listed items.

[0072] Hereinafter, the terms "first" and "second" are only used for descriptive purposes, and cannot be construed as implying or indicating relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present application, unless otherwise stated, the meaning of "a plurality" is two or more.

[0073] The following further describes the embodiments of the present application in conjunction with the accompanying drawings of the specification.

[0074] In a first aspect, the present application provides a method for correcting a microphone audio signal. Referring to Figure 1 , the method includes the following steps:

[0075] S10. Obtain an audio correction mode and target parameters, and obtain overall audio correction parameters according to the audio correction mode and the target parameters.

[0076] In this embodiment, the audio correction mode refers to the correction strategies in different preset scenarios, such as the conference mode, the mobile phone mode, the headphone mode, etc. The target parameters are the expected audio quality indicators, such as the target signal-to-noise ratio, the target voice fidelity, etc. The overall audio correction parameters are a set of system-level parameters used to guide the correction operations of all subsequent channels.

[0077] Specifically, the system provides multiple audio correction mode options for the user to select, or automatically analyzes and determines the applicable mode according to the current scenario. Then, key indicators are extracted from the target parameters as the main correction targets, such as using the target signal-to-noise ratio as the main noise suppression target. According to the selected mode and target indicators, a parameter lookup table is queried to obtain the corresponding noise estimation parameters, adaptive filter parameters, pre-processing parameters, etc., and an overall audio correction parameter set is generated based on the combination relationship of these parameters.

[0078] S20. Obtain the microphone channel information, and set personalized correction parameters for the microphone channels according to the microphone channel information and the overall audio correction parameters.

[0079] In this embodiment, the microphone channel information includes the position information of each microphone channel and the noise characteristic information of this channel; the personalized correction parameters refer to a set of correction parameters for each specific channel, used to guide operations such as noise reduction and feedback suppression for this channel.

[0080] Specifically, by obtaining the microphone channel information, and then combining the overall audio correction parameters according to the microphone channel information, appropriate personalized correction parameters are respectively matched for each channel, and the personalized parameters of different channels will be different.

[0081] S30. Perform audio correction according to the personalized correction parameters and the overall audio correction parameters.

[0082] In this embodiment, the audio correction is a processing process performed on the received original microphone audio signal, including pre-processing, noise reduction, feedback suppression, etc.

[0083] Specifically, for the original audio signal of each channel, first perform pre-processing operations such as pre-emphasis filtering and de-pre-emphasis filtering using the corresponding personalized pre-processing parameters, then perform noise detection and noise suppression, and finally perform feedback suppression processing on the audio signal to eliminate problems such as howling. After all channels complete the above corrections, the final optimized audio output is obtained.

[0084] S40. Perform quality detection on the corrected audio to obtain audio detection parameters; compare the audio detection parameters with the target parameters to obtain the parameter difference value.

[0085] In this embodiment, quality detection refers to analyzing and evaluating the audio output after calibration processing to obtain detection parameters reflecting audio quality, such as the actual signal-to-noise ratio, speech fidelity score, etc. Then, these detection parameters are compared with the pre-set target parameters, and the difference value between the two is calculated.

[0086] Specifically, an audio quality assessment algorithm is used to perform spectral analysis, distortion detection, speech clarity scoring, etc. on the calibrated audio to obtain quality parameters such as the actual signal-to-noise ratio, speech fidelity score, etc. These parameters are compared one by one with the target parameters such as the target signal-to-noise ratio and target speech fidelity set in step S10, and the corresponding difference values are calculated to evaluate whether the current calibration effect meets the standard.

[0087] S50: If the parameter difference value exceeds the preset threshold, trigger audio quality monitoring; according to the audio quality monitoring results, obtain parameter adjustment information.

[0088] Among them, the parameter difference value exceeding the threshold means that there is a large deviation between the current calibration effect and the expectation. At this time, it is necessary to start the audio quality monitoring mechanism to deeply analyze the reasons for the quality decline. According to the monitoring and analysis results, obtain the parameter information that needs to be adjusted, providing a basis for the next parameter update.

[0089] Specifically, this application sets a difference threshold range. For example, a signal-to-noise ratio difference within 3 dB and a speech fidelity score difference within 0.2 points are considered acceptable. If it is detected that the difference value exceeds the threshold range, it is considered that the audio quality is poor, and the system automatically triggers the audio quality monitoring process, including: detecting the current ambient noise situation, checking the microphone array status, analyzing whether the speech mode has changed, etc., and according to the type of problem found in the monitoring, query the problem-parameter mapping table to obtain the parameter items that need to be adjusted and the adjustment methods, and output them as parameter adjustment information.

[0090] S60: Update the personalized calibration parameters according to the parameter adjustment information.

[0091] Specifically, parse out the parameter items that need to be adjusted from the parameter adjustment information, such as needing to increase the noise suppression intensity of a certain channel, increase the feedback suppression coefficient of a certain channel, etc. Then, according to the adjustment method, make corresponding adjustments to the original personalized noise suppression parameters, feedback suppression parameters, etc., to obtain the updated set of personalized calibration parameters.

[0092] In one embodiment, referring to Figure 2 , in step S10, obtain the audio calibration mode and target parameters, and according to the audio calibration mode and target parameters, obtain the overall audio calibration parameters, which specifically include the following steps:

[0093] S21: Extract the target signal-to-noise ratio from the target parameters and use the target signal-to-noise ratio as the main noise suppression target.

[0094] Among them, the target signal-to-noise ratio refers to the signal-to-noise ratio level that the expected final audio output should achieve. In this embodiment, the target signal-to-noise ratio is used as the main noise suppression target and the main optimization direction for calibration.

[0095] Specifically, the parameter of the target signal-to-noise ratio is separately extracted from the set of target parameters set by the user or the system default. If the user does not make a special setting, the corresponding recommended target signal-to-noise ratio value can be found according to the current audio calibration mode (such as the conference mode) and used as the system default value.

[0096] S22. Obtain the noise estimation parameter, the adaptive filter parameter, and the preprocessing parameter according to the target signal-to-noise ratio and the audio calibration mode.

[0097] In this embodiment, effective noise reduction and echo cancellation are specifically achieved through the noise estimation parameter, the adaptive filter parameter, and the preprocessing parameter. For different target signal-to-noise ratios and calibration modes, the setting requirements for the noise estimation parameter, the adaptive filter parameter, and the preprocessing parameter are different.

[0098] Specifically, in this embodiment, a look-up table is established in advance, which stores the corresponding parameter configuration schemes such as the type of noise estimation algorithm, the algorithm parameter, the type of adaptive filter, the filter length, and the type of preprocessing filter according to different combinations of the target signal-to-noise ratio and the audio mode. After determining the target signal-to-noise ratio and the current calibration mode, a search is performed in the look-up table to obtain the set of noise estimation parameters, the set of adaptive filter parameters, and the set of preprocessing parameters that match them.

[0099] S23. Obtain the overall audio calibration parameter according to the association of the target signal-to-noise ratio, the noise estimation parameter, the adaptive filter parameter, and the preprocessing parameter.

[0100] Specifically, according to the target signal-to-noise ratio, a noise suppression intensity adjustment coefficient is set, and then the noise suppression intensity adjustment coefficient is combined with the noise estimation parameter to obtain specific noise estimation sub-parameters such as the expected noise detection threshold and the smoothing factor. Then, the adjustment coefficient is simultaneously applied to the adaptive filter parameter to obtain specific sub-parameters such as the filter length and the adaptive step size. Finally, the preprocessing parameter is fused with the previous parameter set to generate a comprehensive overall audio calibration parameter set including various sub-parameter items.

[0101] In one embodiment, referring to Figure 3 , in step S22, the preprocessing parameter includes the pre-emphasis filter parameter and the de-pre-emphasis filter parameter, and the noise estimation parameter includes the noise spectrum estimation parameter and the noise tracking factor parameter.

[0102] According to the target signal-to-noise ratio and the audio correction mode, obtain the noise estimation parameters, adaptive filter parameters, and preprocessing parameters. The specific steps are as follows:

[0103] S221. According to the target signal-to-noise ratio and the audio correction mode, obtain the pre-emphasis filter parameters, de-pre-emphasis filter parameters, noise spectrum estimation parameters, noise tracking factor parameters, and adaptive filter parameters.

[0104] Specifically, according to different combinations of the target signal-to-noise ratio and the audio mode, store the corresponding parameter configuration schemes such as the pre-emphasis filter type and order, de-pre-emphasis filter type and order, noise spectrum estimation algorithm type and smoothing factor, noise tracking rate factor, adaptive filter type and length.

[0105] In the above lookup table, retrieve the parameter configuration scheme that matches the input target signal-to-noise ratio and audio mode, and respectively read out the pre-emphasis filter type and its order, de-pre-emphasis filter type and its order from it as the pre-emphasis / depre-emphasis filter parameters; at the same time, read out the noise spectrum estimation algorithm type and its smoothing factor as the noise spectrum estimation parameters; read out the rate factor of noise tracking as the noise tracking factor parameters; read out the adaptive filter type and its length as the adaptive filter parameters.

[0106] S222. According to the noise spectrum estimation parameters and the noise tracking factor parameters, obtain the noise estimation parameters.

[0107] Specifically, according to the noise spectrum estimation parameters and the noise tracking factor parameters, a final set of noise estimation parameters is formed.

[0108] In one embodiment, refer to Figure 4 , in step S30, the personalized correction parameters include the channel noise suppression parameter and the channel feedback suppression parameter. Obtain the microphone channel information, and according to the microphone channel information and the overall audio correction parameters, set the personalized correction parameters for the microphone channel. The specific steps are as follows:

[0109] S31. Obtain the microphone channel position information and the channel noise characteristic information.

[0110] In this embodiment, the microphone channel position information refers to the specific geometric position of the microphone in the array, and the channel noise characteristic information describes the type, spectral distribution, and time duration characteristics of the noise received by this channel, etc.

[0111] Specifically, automatically obtain the three-dimensional coordinate positions of each microphone channel in the microphone array. At the same time, use the noise monitoring unit to detect the background noise of each channel respectively, analyze its spectral power, duration, change rate, etc., and form the noise characteristic data of this channel.

[0112] S32. Obtain channel pre - processing parameters for the microphone channel according to the pre - processing parameters, channel position information, and channel noise characteristic information.

[0113] In this embodiment, the channel pre - processing parameters are used to perform pre - processing such as pre - emphasis and de - pre - emphasis on the original audio signal of the channel to improve the subsequent noise reduction effect. The positions and noise characteristics of different channels are different, and the required pre - processing parameters will also be different.

[0114] Specifically, this application presets a look - up table. According to the channel position and noise type (such as low - frequency noise, high - frequency noise), etc., all possible situations are divided into several grades, and the corresponding pre - emphasis filter type, order, de - pre - emphasis filter type, order and other pre - processing parameter configurations are stored for each grade. After obtaining the channel position and noise information, search for the matching grade in the look - up table, and read out the corresponding set of pre - processing parameters as the personalized pre - processing parameters for this channel. Optionally, the farther the distance from the sound source position, the higher the grade setting; the higher the noise frequency, the higher the grade setting.

[0115] S33. Obtain channel noise suppression parameters and channel feedback suppression parameters for the microphone channel according to the noise estimation parameters, adaptive filter parameters, channel position information, and channel noise characteristic information.

[0116] In this embodiment, the channel noise suppression parameters are used to guide the noise detection and suppression operations of this channel, and the channel feedback suppression parameters are used to guide anti - howling and echo suppression.

[0117] Similarly, a look - up table can be preset. According to different combinations of factors such as channel position and noise type, the corresponding noise detection threshold, smoothing length and other noise suppression sub - parameters, as well as the configuration schemes of feedback suppression sub - parameters such as adaptive filter length and adaptive step size are stored for each situation. After obtaining the channel position and noise information, retrieve the matching scheme in the look - up table, and read out the corresponding set of noise suppression parameters and feedback suppression parameters. At the same time, fuse these parameters with the noise estimation parameters and adaptive filter parameters in the overall parameters respectively to form the final noise suppression and feedback suppression parameters for this channel.

[0118] In one embodiment, referring to Figure 5 , in step S40, according to the audio quality monitoring result, obtain parameter adjustment information, which specifically includes the following steps:

[0119] S41. Obtain audio quality change information according to the audio quality monitoring result.

[0120] In this embodiment, the audio quality change information reflects the change trend of the current audio quality relative to the expected target, such as continuous quality decline, sudden deterioration, or fluctuations.

[0121] Specifically, this embodiment continuously tracks key metrics in the audio quality monitoring results, such as real-time signal-to-noise ratio, speech fidelity score, etc., and compares them with previous historical values. If it is detected that these metrics show a continuous downward trend, it is determined that the audio quality is continuously deteriorating; if there is a sudden large quality distortion, it is determined that the audio quality suddenly deteriorates; if the metric values fluctuate greatly, it is determined that the audio quality is unstable, etc.

[0122] S42. Obtain quality anomaly trigger information, and trigger an environment detection instruction according to the quality anomaly trigger information.

[0123] In this embodiment, the quality anomaly trigger information refers to some criterion conditions pre-set inside the system, which are used to judge whether the current audio quality has become abnormal to the extent that the environment needs to be detected and diagnosed. Once it is detected that the audio quality change information matches these conditions, the system will automatically trigger relevant instructions for environment detection.

[0124] Specifically, set quality anomaly trigger conditions such as "the duration of continuous quality deterioration exceeds 5 minutes", "the amplitude of sudden quality deterioration exceeds 5 dB", "the amplitude of quality fluctuation exceeds 3 dB and the duration exceeds 2 minutes", etc. These conditions can be set in advance according to experience, or can be adjusted autonomously during the operation of the system. When it is monitored that the metric values of audio quality change information such as the decline amplitude of real-time signal-to-noise ratio and the decline degree of speech fidelity score meet these conditions, the system will determine that the audio quality is abnormal and immediately trigger an environment detection instruction.

[0125] S43. Obtain environment detection result information, and obtain the current audio processing status information according to the environment detection result information in combination with the audio quality change information.

[0126] In this embodiment, the current audio processing status information reflects the specific reasons for the decline in audio quality, such as environmental noise change, microphone failure, or voice mode switching, etc.

[0127] Specifically, judge whether the noise exceeds the system's adaptation ability through environment detection; at the same time, check whether the operating status of each microphone channel in the microphone array is normal. Combine these detection results with the previously obtained audio quality change information (such as continuous quality decline, sudden deterioration, or fluctuating instability, etc.) for comprehensive analysis to judge the current audio processing status, such as being affected by a certain type of sudden noise, due to the failure of a certain microphone channel, or the voice mode changing from a conference to a video call, etc.

[0128] S44. Obtain parameter adjustment information according to the current audio processing status information and the overall audio correction parameters.

[0129] Specifically, according to the current processing status information obtained in the previous step, search for a parameter adjustment strategy that matches it. For example, when encountering a certain type of low-frequency noise, increase the smoothing length in the noise suppression parameters of some channels and lower the detection threshold, etc.; when a certain microphone channel fails, increase the weight of other channels and appropriately adjust the adaptive filter parameters, etc.

[0130] In one embodiment, referring to Figure 6 , in step S60, according to the parameter adjustment information, update the personalized correction parameters, which specifically includes the following steps:

[0131] S61. According to the parameter adjustment information, obtain the personalized parameter adjustment information.

[0132] Among them, the personalized parameter adjustment information includes the adjusted channel noise suppression parameters and the adjusted channel feedback suppression parameters.

[0133] Specifically, by traversing all the adjustment items included in the parameter adjustment information, classify the channel noise suppression sub-parameters (such as smoothing length, detection threshold, etc.) and feedback suppression sub-parameters (such as adaptive filter length, adaptive step size, etc.) involved according to the channels to form a personalized adjustment subset for each channel.

[0134] S62. Compare the adjusted channel noise suppression parameters and the adjusted channel feedback suppression parameters with the preset parameter thresholds. If they are within the threshold range, update the personalized correction parameters according to the personalized parameter adjustment information.

[0135] In this embodiment, to prevent the over-adjustment of personalized parameters from affecting the system stability, limit the adjustment amplitude of each channel. Only when the adjusted parameter value is within the reasonable threshold range will the actual update be executed.

[0136] Specifically, a change threshold range can be set in advance for each sub-parameter of channel noise suppression and feedback suppression, such as the noise smoothing length is in [5, 50], the adaptive filter length is in [64, 512], etc. After obtaining the personalized adjustment information of this channel, compare the adjusted parameter value with the corresponding threshold range. If the adjusted value is still within the threshold range, directly update the personalized correction parameters of this channel according to the adjustment information; if the adjusted value exceeds the threshold range, temporarily do not execute the update.

[0137] S63. If it exceeds the threshold range, trigger a parameter secondary adjustment instruction. According to the parameter secondary adjustment instruction, obtain the secondary adjustment parameter information, and update the personalized correction parameters according to the secondary adjustment parameter information.

[0138] Specifically, after detecting that the adjustment amplitude of a certain channel exceeds the threshold, an instruction for secondary adjustment of the channel parameters is automatically triggered; the strategy for secondary adjustment can be to apply a reduction coefficient to the initial adjustment amplitude to reduce the adjustment amplitude to an acceptable range. For example, if the noise smoothing length of a certain channel is initially adjusted from 20 to 60, which exceeds the range of [5, 50], the secondary adjustment can reduce 60 to 40, falling within the threshold range. After obtaining the information of the secondary adjustment parameters after reduction, the original personalized parameter adjustment information is replaced with this information, and finally the actual update of the personalized correction parameters of the channel is completed, so as to balance the effective adjustment of the parameters and the stability of the system.

[0139] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The order of execution of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0140] In a second aspect, the present application provides a microphone audio signal correction system. The microphone audio signal correction system of the present application will be described below in combination with the above microphone audio signal correction method.

[0141] Referring to Figure 7 , a microphone audio signal correction system includes:

[0142] An audio correction mode acquisition module, configured to acquire an audio correction mode;

[0143] A target parameter acquisition module, configured to acquire target parameters;

[0144] An overall parameter generation module, configured to acquire overall audio correction parameters according to the audio correction mode and the target parameters;

[0145] A channel parameter setting module, configured to acquire microphone channel information, and set personalized correction parameters for each microphone channel according to the microphone channel information and the overall audio correction parameters;

[0146] An audio correction module, configured to perform audio correction according to the personalized correction parameters and the overall audio correction parameters;

[0147] A quality detection module, configured to perform quality detection on the corrected audio to acquire audio detection parameters;

[0148] A difference acquisition module, configured to compare the audio detection parameters with the target parameters to obtain a parameter difference value;

[0149] An adaptive adjustment trigger module, configured to trigger audio quality monitoring when the parameter difference value exceeds a preset threshold;

[0150] A parameter adjustment module, configured to acquire parameter adjustment information according to an adaptive adjustment mechanism;

[0151] A parameter update module, configured to update personalized correction parameters according to parameter adjustment information.

[0152] In an optional embodiment, the overall parameter generation module includes:

[0153] A target extraction unit, configured to extract a target signal-to-noise ratio from target parameters and use the target signal-to-noise ratio as the main noise suppression target;

[0154] A correction parameter acquisition unit, configured to acquire noise estimation parameters, adaptive filter parameters, and preprocessing parameters according to the target signal-to-noise ratio and the audio correction mode;

[0155] An overall parameter generation unit, configured to obtain overall audio correction parameters by associating the target signal-to-noise ratio, noise estimation parameters, adaptive filter parameters, and preprocessing parameters.

[0156] In an optional embodiment, the preprocessing parameters include pre-emphasis filter parameters and de-pre-emphasis filter parameters, and the noise estimation parameters include noise spectrum estimation parameters and noise tracking factor parameters;

[0157] The correction parameter acquisition unit includes:

[0158] A preprocessing parameter acquisition unit, configured to acquire pre-emphasis filter parameters and de-pre-emphasis filter parameters according to the target signal-to-noise ratio and the audio correction mode;

[0159] A noise estimation parameter acquisition unit, configured to acquire noise spectrum estimation parameters, noise tracking factor parameters, and acquire adaptive filter parameters according to the target signal-to-noise ratio and the audio correction mode;

[0160] A noise estimation generation unit, configured to acquire noise estimation parameters according to the noise spectrum estimation parameters and the noise tracking factor parameters.

[0161] In an optional embodiment, the personalized correction parameters include channel noise suppression parameters and channel feedback suppression parameters;

[0162] The channel parameter setting module includes:

[0163] A channel information acquisition unit, configured to acquire microphone channel position information and channel noise feature information;

[0164] A channel preprocessing parameter acquisition unit, configured to acquire channel preprocessing parameters for each microphone channel according to the preprocessing parameters, channel position information, and channel noise feature information;

[0165] A channel noise reduction feedback parameter acquisition unit, which is used to acquire channel noise suppression parameters and channel feedback suppression parameters for each microphone channel according to noise estimation parameters, adaptive filter parameters, channel position information, and channel noise characteristic information.

[0166] In an optional embodiment, the parameter adjustment module includes:

[0167] An audio quality change acquisition unit, which is used to acquire audio quality change information according to the audio quality monitoring result;

[0168] A quality convergence trigger unit, which is used to acquire quality anomaly trigger information;

[0169] An environment detection trigger unit, which is used to trigger an environment detection instruction according to the quality anomaly trigger information;

[0170] An environment detection result acquisition unit, which is used to acquire environment detection result information;

[0171] An audio state acquisition unit, which is used to acquire current audio processing state information according to the environment detection result information and in combination with the audio quality change information;

[0172] A parameter adjustment information acquisition unit, which is used to acquire parameter adjustment information according to the current audio processing state information and the overall audio correction parameters.

[0173] In an optional embodiment, the parameter update module includes:

[0174] A personalized parameter adjustment unit, which is used to acquire personalized parameter adjustment information according to the parameter adjustment information, where the personalized parameter adjustment information includes the adjusted channel noise suppression parameters and the adjusted channel feedback suppression parameters;

[0175] A parameter threshold comparison unit, which is used to compare the adjusted channel noise suppression parameters and the adjusted channel feedback suppression parameters with preset parameter thresholds;

[0176] A parameter update unit, which is used to update the personalized correction parameters according to the personalized parameter adjustment information when the adjusted parameters are within the threshold range;

[0177] A secondary adjustment trigger unit, which is used to trigger a parameter secondary adjustment instruction when the adjusted parameters exceed the threshold range;

[0178] A secondary adjustment parameter acquisition unit, which is used to acquire secondary adjustment parameter information according to the parameter secondary adjustment instruction;

[0179] The parameter update unit is further used to update the personalized correction parameters according to the secondary adjustment parameter information.

[0180] In one embodiment, the present application provides an electronic device, which may be a server, and its internal structure diagram may be as shown in Figure 8 . The electronic device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the electronic device is used to store data. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a method for correcting microphone audio signals.

[0181] Those skilled in the art can understand that Figure 8 the structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the electronic device to which the solution of the present application is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0182] In one embodiment, an electronic device is further provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0183] Those of ordinary skill in the art can understand that all or part of the processes of implementing the above method embodiments can be completed by instructing relevant hardware through a computer program. The above computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the above method embodiments. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided by the present application may include at least one of non-volatile and volatile memories. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical memory, etc. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0184] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereby. Therefore, all equivalent changes made according to the structure, shape, and principle of the present application shall be covered within the protection scope of the present application.

Claims

1. A microphone audio signal correction method, characterized in that: The steps include: Obtain an audio correction mode and a target parameter, and obtain an overall audio correction parameter according to the audio correction mode and the target parameter, wherein the audio correction mode refers to a correction strategy for different preset scenarios, including a conference mode, a mobile phone mode, and a headset mode; Acquire microphone channel information, and set personalized correction parameters for the microphone channel according to the microphone channel information and the overall audio correction parameters, where the personalized correction parameters refer to a correction parameter set for each specific channel; Performing audio correction according to the personalized correction parameter and the overall audio correction parameter; Performing quality detection on the corrected audio to obtain audio detection parameters; comparing the audio detection parameters with the target parameters to obtain parameter difference values; If the parameter difference value exceeds a preset threshold, audio quality monitoring is triggered; and parameter adjustment information is obtained according to the audio quality monitoring result; updating the personalized correction parameters according to the parameter adjustment information; The step of obtaining the audio correction mode and the target parameters, and obtaining the overall audio correction parameters according to the audio correction mode and the target parameters, specifically includes the following steps: Extracting a target signal-to-noise ratio from the target parameters, and using the target signal-to-noise ratio as a main noise suppression target; Acquire noise estimation parameters, adaptive filter parameters, and pre-processing parameters according to the target signal-to-noise ratio and the audio correction mode; Obtaining the overall audio correction parameters according to the association of the target signal-to-noise ratio, the noise estimation parameters, the adaptive filter parameters and the pre-processing parameters; The acquiring of microphone channel information and setting of personalized correction parameters for the microphone channel according to the microphone channel information and the overall audio correction parameters specifically includes the following steps: Obtain microphone channel position information and channel noise characteristic information; Acquire channel pre-processing parameters for the microphone channel according to the pre-processing parameters, the channel position information and the channel noise characteristic information; A channel noise suppression parameter and a channel feedback suppression parameter are obtained for the microphone channel according to the noise estimation parameter, the adaptive filter parameter, the channel position information and the channel noise characteristic information.

2. The microphone audio signal correction method according to claim 1, characterized in that: The pre-processing parameters include pre-emphasis filter parameters and de-pre-emphasis filter parameters, and the noise estimation parameters include noise spectrum estimation parameters and noise tracking factor parameters. The step of acquiring noise estimation parameters, adaptive filter parameters and pre-processing parameters according to the target signal-to-noise ratio and the audio correction mode specifically comprises the following steps: According to the target signal-to-noise ratio and the audio correction mode, acquiring the pre-emphasis filter parameters, the de-pre-emphasis filter parameters, the noise spectrum estimation parameters, the noise tracking factor parameters and the adaptive filter parameters; The noise estimation parameter is acquired according to the noise spectrum estimation parameter and the noise tracking factor parameter.

3. The microphone audio signal correction method according to claim 1, characterized in that: The step of obtaining parameter adjustment information according to the audio quality monitoring result specifically includes the following steps: Acquiring audio quality change information according to the audio quality monitoring result; Acquire quality abnormality trigger information, and trigger an environmental detection instruction according to the quality abnormality trigger information; Acquire environment detection result information, and acquire current audio processing state information according to the environment detection result information and the audio quality change information; Parameter adjustment information is obtained according to the current audio processing state information and the overall audio correction parameter.

4. The microphone audio signal correction method according to claim 1, characterized in that: The updating of the personalized correction parameters according to the parameter adjustment information specifically includes the following steps: According to the parameter adjustment information, obtaining personalized parameter adjustment information, wherein the personalized parameter adjustment information includes an adjusted channel noise suppression parameter and an adjusted channel feedback suppression parameter; Comparing the adjusted channel noise suppression parameter and the adjusted channel feedback suppression parameter with a preset parameter threshold, and if they are within the threshold range, updating the personalized correction parameter according to the personalized parameter adjustment information; If it exceeds the threshold range, a parameter secondary adjustment instruction is triggered, and secondary adjustment parameter information is obtained according to the parameter secondary adjustment instruction, and the personalized correction parameter is updated according to the secondary adjustment parameter information.

5. A microphone audio signal correction system, characterized in that: include: An audio correction mode acquisition module is used to acquire an audio correction mode, where the audio correction mode refers to a correction strategy for different preset scenarios, including a conference mode, a mobile phone mode, and a headset mode; A target parameter acquisition module is used to obtain target parameters; An overall parameter generation module, used for acquiring overall audio correction parameters according to the audio correction mode and the target parameters; A channel parameter setting module, used to obtain microphone channel information, and set personalized correction parameters for each microphone channel according to the microphone channel information and the overall audio correction parameters, wherein the personalized correction parameters refer to a correction parameter set for each specific channel; An audio correction module, used for performing audio correction according to the personalized correction parameters and the overall audio correction parameters; A quality detection module is used to perform quality detection on the corrected audio and obtain audio detection parameters; A difference acquisition module, used for comparing the audio detection parameter with the target parameter to obtain a parameter difference value; An adaptive adjustment trigger module, used to trigger audio quality monitoring when the parameter difference value exceeds a preset threshold; A parameter adjustment module, used for obtaining parameter adjustment information according to an adaptive adjustment mechanism; A parameter updating module, used for updating the personalized correction parameters according to the parameter adjustment information; The parameter updating module comprises: A personalized parameter adjustment unit, configured to obtain personalized parameter adjustment information according to the parameter adjustment information, wherein the personalized parameter adjustment information includes an adjusted channel noise suppression parameter and an adjusted channel feedback suppression parameter; A parameter threshold comparison unit, used for comparing the adjusted channel noise suppression parameter and the adjusted channel feedback suppression parameter with a preset parameter threshold; a parameter updating unit, configured to update the personalized correction parameter according to the personalized parameter adjustment information when the adjusted parameter is within a threshold range; A secondary adjustment trigger unit, used for triggering a parameter secondary adjustment instruction when the adjusted parameter exceeds a threshold range; A secondary adjustment parameter acquisition unit, used to acquire secondary adjustment parameter information according to the parameter secondary adjustment instruction; The parameter updating unit is further used to update the personalized correction parameter according to the secondary adjustment parameter information; The channel parameter setting module includes: A channel information acquisition unit, used to acquire microphone channel position information and channel noise characteristic information; a channel pre-processing parameter acquisition unit, configured to acquire a channel pre-processing parameter for each microphone channel according to the pre-processing parameter, the channel position information and the channel noise characteristic information; A channel noise reduction feedback parameter acquisition unit is used to acquire a channel noise suppression parameter and a channel feedback suppression parameter for each microphone channel according to the noise estimation parameter, the adaptive filter parameter, the channel position information and the channel noise characteristic information.

6. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the microphone audio signal correction method according to any one of claims 1 to 4 when executing the computer program.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the microphone audio signal correction method according to any one of claims 1 to 4 are implemented.

Images