A wireless audio sound effect processing method and system

Abstract

The present invention provides a wireless audio sound effect processing method and system, relating to the technical field of audio processing. The method includes: collecting wireless audio data; determining whether there is noise in the wireless audio data through an autoregressive prediction algorithm, and if there is, extracting the noise signal in the wireless audio data; combining a sparse dictionary learning algorithm, and generating a reverse noise signal for canceling the noise signal through a filter with an adaptive filtering step size; aligning the noise signal and the reverse noise signal based on a dynamic phase compensation order; synthesizing the wireless audio data and the aligned reverse noise signal to obtain optimized wireless audio data; otherwise, marking the wireless audio data as optimized wireless audio data and outputting the optimized wireless audio data. While improving the noise reduction efficiency, it can effectively cope with complex and variable noise information and improve the quality of wireless audio.

Description

Technical Field

[0001] The present invention relates to the technical field of audio processing, and particularly to a wireless audio sound effect processing method and system. Background Art

[0002] With the rapid development of human-computer interaction and wireless devices, currently, both microphones and Bluetooth headsets adopt noise reduction technologies to improve audio collection effects or audio playback effects. The noise reduction technology can effectively remove background noise, improve speech clarity, optimize the sound quality experience, and ensure high-quality sound transmission even in a noisy environment.

[0003] Audio noise reduction is crucial for improving audio quality. Especially in modern communication, speech recognition, broadcasting, and entertainment applications, noise can greatly interfere with information transmission. Noise not only affects the user's auditory experience but also may cause signal distortion, reducing the accuracy and efficiency of the system. Through noise reduction technology, the effective information in the audio signal can be clearly extracted, enhancing the reliability and accuracy of the speech recognition system, while improving the listening quality and avoiding the interference of background noise on communication.

[0004] However, existing noise reduction methods often use fixed noise reduction parameters to perform audio noise reduction on different complex noise environments, which may lead to sound quality loss, especially in the case of transient noise or high-frequency noise, the effect is poor. Summary of the Invention

[0005] In order to solve the technical problem that existing noise reduction methods often use fixed noise reduction parameters to perform audio noise reduction on different complex noise environments, which may lead to sound quality loss, especially in the case of transient noise or high-frequency noise, the effect is poor, the present invention provides a wireless audio sound effect processing method and system.

[0006] The technical solutions provided by the embodiments of the present invention are as follows: [[ID=2�]]

[0007] First Aspect

[0008] A wireless audio sound effect processing method provided by an embodiment of the present invention includes:

[0009] S1: Collect wireless audio data;

[0010] S2: Determine whether there is noise in the wireless audio data through an autoregressive prediction algorithm. If there is noise, enter step S3; otherwise, mark the wireless audio data as optimized wireless audio data and enter step S7;

[0011] S3: Extract the noise signal in the wireless audio data;

[0012] S4: Combine the sparse dictionary learning algorithm to generate a reverse noise signal for canceling the noise signal through a filter with an adaptive filtering step size;

[0013] S5: Align the noise signal and the reverse noise signal based on the dynamic phase compensation order;

[0014] S6: Synthesize the wireless audio data and the aligned reverse noise signal to obtain optimized wireless audio data;

[0015] S7: Output the optimized wireless audio data.

[0016] Second aspect

[0017] A wireless audio sound effect processing system provided by an embodiment of the present invention includes:

[0018] A processor;

[0019] A memory, on which computer-readable instructions are stored. When the computer-readable instructions are executed by the processor, the wireless audio sound effect processing method described in the first aspect is implemented.

[0020] Third aspect

[0021] A computer-readable storage medium provided by an embodiment of the present invention, on which a computer program is stored. When the program is executed by a processor, the wireless audio sound effect processing method described in the first aspect is implemented.

[0022] The beneficial effects brought by the technical solution provided by the embodiment of the present invention at least include:

[0023] In the embodiments of the present invention, an autoregressive prediction algorithm is first used to perform pre-noise judgment on the collected wireless audio data. Only when noise is judged to exist can noise reduction be carried out. By analyzing the autocorrelation characteristics of the audio data, the presence or absence of noise can be accurately judged, avoiding noise reduction in the absence of noise, thereby improving the noise reduction efficiency and audio quality, effectively avoiding unnecessary noise reduction processing, enhancing the computing efficiency, reducing the audio loss caused by global noise reduction, and at the same time improving the accuracy of noise judgment. During the noise reduction process, a sparse dictionary learning algorithm is combined. A reverse noise signal for canceling the noise signal is generated by a filter with an adaptive filtering step size. The sparse dictionary learning algorithm can effectively extract the noise components by learning the noise characteristics and performing sparse representation, while the adaptive filtering step size dynamically adjusts the filter according to the noise change, thereby achieving more accurate noise cancellation and better audio quality. And the noise signal and the reverse noise signal are aligned based on the dynamic phase compensation order, and then the wireless audio data and the aligned reverse noise signal are synthesized to obtain optimized wireless audio data and output. It can accurately judge noise, dynamically adjust the filter, and extract noise characteristics, thereby effectively improving the accuracy of noise cancellation, avoiding unnecessary noise reduction processing, and at the same time enhancing the audio quality and processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0025] Figure 1 FIG. is a schematic flowchart of a method for processing wireless audio sound effects provided by an embodiment of the present invention;

[0026] Figure 2 FIG. is a schematic structural diagram of a system for processing wireless audio sound effects provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] The following describes the technical solutions in the present invention with reference to the drawings.

[0028] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations, or explanations. Any embodiment or design solution described as "example" in the present invention should not be construed as being more excellent or having more advantages than other embodiments or design solutions. Exactly speaking, the use of the word "example" is intended to present concepts in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0029] To make the technical problems, technical solutions, and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0030] Refer to the attached Figure 1 FIG. shows a schematic flowchart of a wireless audio sound effect processing method provided by an embodiment of the present invention.

[0031] An embodiment of the present invention provides a wireless audio sound effect processing method, which can be implemented by a wireless audio sound effect processing system device, and the wireless audio sound effect processing system device can be a terminal or a server. The processing flow of the wireless audio sound effect processing method may include the following steps:

[0032] S1: Collect wireless audio data.

[0033] Among them, wireless audio data refers to an audio signal transmitted through a wireless signal, which does not rely on a wired connection and is usually transmitted from an audio source (such as a microphone, earphone) to a receiving device through a wireless communication technology such as Bluetooth, Wi-Fi or other wireless protocols. The wireless audio data may include audio acquisition data from a microphone, or may also be audio data collected by an earphone through a built-in microphone, including the output audio signal of the earphone and the reverse audio signal for noise reduction. The wireless audio data is collected through a wireless device (such as a microphone or a Bluetooth earphone). These data include the audio signal received from the environment and the sound collected by the built-in microphone of the earphone, which is used to achieve more efficient noise reduction. The collected audio data will be used as the input for subsequent noise reduction processing to ensure that the noise signal can be effectively identified and removed.

[0034] S2: Determine whether there is noise in the wireless audio data through an autoregressive prediction algorithm. If there is, go to step S3; otherwise, mark the wireless audio data as optimized wireless audio data and go to step S7.

[0035] Among them, the autoregressive prediction algorithm is a statistical model that predicts future values through current and past signal values. The autoregressive model assumes that the value of the current signal is a linear combination of the signals in the previous period. By estimating the autoregressive coefficients, the current signal can be predicted based on historical data, thereby identifying the noise components in the signal. The wireless audio data is analyzed through the autoregressive prediction algorithm to determine whether there is noise in the signal. The algorithm predicts the future signal value by analyzing the autocorrelation characteristics of the signal and compares it with the actual value to identify the noise. If noise is detected, the system proceeds to the next step for noise reduction processing. If there is no noise, the wireless audio data is directly marked as optimized data and jumps to the subsequent steps to ensure efficiency and audio quality.

[0036] In a possible implementation manner, S201: Frame the wireless audio data in combination with the autoregressive order related to the maximum noise length:

[0037]

[0038] Among them, N w represents the frame length, p represents the autoregressive order of the autoregressive prediction algorithm, and N max represents the maximum noise length, and N h represents the overlap degree between frames.

[0039] Among them, the maximum noise length refers to the longest time or the maximum number of samples that the noise signal may last in the wireless audio signal. It helps to determine the order of the autoregressive prediction model to better capture the characteristics of the noise. By combining the maximum noise length to determine the order of the autoregressive prediction algorithm, and then determining the frame length and the overlap degree between frames. In this way, the audio signal can be accurately divided and the noise-related characteristics can be extracted, ensuring the accuracy of noise judgment and the effectiveness of noise reduction processing.

[0040] S202: Calculate the autocorrelation coefficients of each frame of wireless audio data obtained by framing at different lag orders and the noise variance related to the prediction error of the autoregressive prediction algorithm:

[0041]

[0042] Among them, represents the estimated value of the autocorrelation function of the wireless audio data at the lag coefficient τ for each frame of wireless audio data, N represents the total number of samples in the wireless audio data of the current frame, and x k and x k-τ represent the k-th wireless audio data value and the (k - τ)-th wireless audio data value in the wireless audio data of the current frame respectively, and a i represents the autocorrelation coefficient of the wireless audio data of the current frame at the i-th order, i = 1, 2, p, represents the noise variance, and e k represents the prediction error of the k-th wireless audio data value in the wireless audio data of the current frame.

[0043] It should be noted that by calculating the autoregressive coefficients and the noise variance, the characteristics of the audio data, especially the behavior of the noise, can be accurately evaluated and modeled. Through the autocorrelation function estimation and the noise variance of the prediction error, the model parameters can be dynamically adjusted to optimize the noise identification and removal effect, ensuring that the noise reduction process is efficient and has a low error, while avoiding unnecessary signal loss and improving the accuracy of noise reduction and the audio quality.

[0044] S203: Combine the calculated autocorrelation coefficients to calculate the difference between the actual value and the predicted value of the wireless audio data of the detection signal:

[0045]

[0046] Among them, represents the actual value of the wireless audio data at time t of the current frame, represents the wireless audio data value at a lag of i, represents the detection signal difference at time t.

[0047] S204: According to the detection signal difference, use the noise standard deviation to determine whether there is noise in the wireless audio data:

[0048]

[0049] Among them, I represents an exponential variable. When I = 1, it means there is noise in the wireless audio data. When I = 0, it means there is no noise in the wireless audio data, represents the noise standard deviation, represents the noise screening threshold.

[0050] Specifically, this process uses the autoregressive (AR) prediction algorithm to judge and analyze the noise of the wireless audio data. First, calculate the autoregressive order and the frame length according to the maximum noise length, and divide the audio data into multiple frames. Then, calculate the autocorrelation coefficients of each frame of audio data at different lag orders, and estimate the noise variance. Through the analysis of the autoregressive coefficients and the noise variance, obtain the prediction error of the signal. By calculating the detection signal difference between the actual value and the predicted value of each frame, further extract the noise characteristics. Finally, based on the detection signal difference and the noise standard deviation, judge whether there is noise in the signal by setting the noise screening threshold. If the difference exceeds the threshold, the system will consider that this frame contains noise and enter the subsequent noise reduction process, otherwise mark it as optimized audio data. By using the autoregressive (AR) prediction algorithm for noise judgment and extraction, the noise components in the wireless audio can be efficiently identified. By calculating the autoregressive coefficients and the noise variance, the noise and effective information in the signal can be accurately distinguished, thus avoiding unnecessary noise reduction processing and improving the processing efficiency. In addition, using the noise standard deviation and the threshold to judge the existence of noise helps to improve the accuracy of noise recognition, reduce the distortion of the audio signal, and ensure optimized audio output without affecting the sound quality.

[0051] Optionally, the noise screening threshold can be specifically set to .

[0052] S3: Extract the noise signal from the wireless audio data.

[0053] It should be noted that by analyzing the difference of the detection signals calculated by the autoregressive prediction algorithm, the noise signals in the wireless audio data are extracted. The noise signals refer to the parts that are inconsistent with the original audio signals, and these parts are regarded as background noise or interference. Through this method, the noise can be effectively separated and prepared for subsequent noise cancellation processing, thereby ensuring the optimized audio quality.

[0054] In a possible implementation manner, S3 is specifically as follows:

[0055] Extract the difference of the detection signals as the noise signals.

[0056] It should be noted that extracting the difference of the detection signals as the noise signals can accurately distinguish the noise from the valid signals. By directly extracting the noise signals, the misprocessing of the original audio is avoided, and the efficiency of noise removal and the audio quality are improved.

[0057] S4: Combine the sparse dictionary learning algorithm, and generate a reverse noise signal for canceling the noise signals through a filter with an adaptive filtering step size.

[0058] Among them, the sparse dictionary learning algorithm is a signal processing technology. By extracting the dictionary (i.e., basis vectors) from the noise data and learning the sparse representation, it aims to represent the features of the signal with fewer non-zero coefficients. This method can effectively separate the noise from the complex signals. The adaptive filtering step size refers to that the filter dynamically adjusts the step size value according to the change of the noise signals. By selecting an appropriate step size, while removing the noise, the quality of the original audio signal can be maximally retained. The filter is a tool for processing signals, used to filter out the noise within a specific frequency range. The adaptive filter can dynamically adjust its coefficients according to the characteristics of the input signal to achieve the best noise cancellation effect. The reverse noise signal is a signal generated by the filter and opposite to the noise signal, and its purpose is to cancel the noise components in the original audio by adding. Through phase alignment and amplitude control, the reverse noise signal can effectively cancel the noise components in the original audio.

[0059] It should be noted that by combining the sparse dictionary learning algorithm and the adaptive filtering step size, the system learns the sparse features of the noise signals and uses the filter to generate the reverse noise signal. The sparse dictionary learning enables the noise components to be accurately extracted and represented, while the adaptive filtering step size adjusts the filter according to the dynamic change of the noise signals, thereby generating a reverse signal that can effectively cancel the noise. This process ensures more efficient noise cancellation while minimizing the loss of sound quality.

[0060] In a possible implementation manner, S4 specifically includes:

[0061] S401: Decompose the noise signals into multiple sub-bands.

[0062] Optionally, wavelet transform can be used for decomposition.

[0063] S402: Perform sparse dictionary learning on each sub-band to obtain the sparse coefficients corresponding to the noise characteristics:

[0064]

[0065] Among them, represents the th sub-band obtained by decomposing the noise signal, represents corresponding dictionary matrix, represents corresponding sparse coefficient vector, represents the regularization parameter representing the sparsity constraint weight, represents the square of the L2 norm, represents the L1 norm, represents taking and when the function is minimized.

[0066] It should be noted that by decomposing the noise signal into multiple sub-bands, the noise characteristics of each band can be processed independently, thereby improving the accuracy of noise suppression. The sparse dictionary learning algorithm can extract the sparse characteristics of the noise from each sub-band and optimize the sparse coefficients through the L1 norm and L2 norm to ensure more accurate noise labeling.

[0067] Optionally, . Specifically, the sparse coefficient vector can be learned through the K-SVD algorithm. Specifically, the steps of learning the sparse coefficient vector through the K-SVD algorithm include: first, initialize the dictionary matrix, and in each iteration, calculate the coefficient vector through sparse representation according to the current dictionary matrix and the noise signal. Then, fix the coefficient vector and update the dictionary matrix using singular value decomposition (SVD) to make the dictionary more adaptable to the sparse representation of the signal. This process gradually improves the accuracy and effectiveness of noise labeling by alternately optimizing the dictionary and sparse coefficients until the preset optimization goal is reached.

[0068] S403: Determine the adaptive filtering step size of the filter according to the sparse coefficient vector:

[0069] ;

[0070] Among them, represents the number of non-zero elements in the sparse coefficient vector, Z represents the number of dictionary atoms, and represent the maximum filtering step size and the minimum filtering step size respectively, A constant representing the avoidance of a zero denominator. In the case where the noise signal is transient noise, the adaptive filtering step size is calculated as follows , and in the case where the noise signal is steady-state noise, the adaptive filtering step size is calculated as follows .

[0071] Among them, the number of dictionary atoms is usually default set to 256. = 10 -6 . The maximum filtering step size can be set to 0.1, and the minimum filtering step size can be set to 0.001. Transient noise and steady-state noise can be judged by whether the noise change rate exceeds a preset noise change rate. If it exceeds the preset noise change rate, where and respectively represent the sub-bands at time t and time t - 1 , then it is considered transient noise, otherwise, it is considered steady-state noise, and then the corresponding adaptive filtering step size is selected. It can be understood that those skilled in the art can set the size of the preset noise change rate according to actual needs, and the present invention does not limit this here.

[0072] It should be noted that by dynamically adjusting the step size of the filter according to the number of non-zero elements of the sparse coefficient vector and the energy of the noise signal, it can flexibly handle transient noise and steady-state noise. By setting the maximum and minimum step sizes, as well as the judgment of the noise change rate, the filter can optimize the filtering effect under different noise conditions, effectively eliminate noise, and prevent over-filtering, thus ensuring the balance between audio quality and processing efficiency.

[0073] In a possible implementation manner, when the noise change rate is greater than the preset noise change rate, it is determined that the noise signal is transient noise, otherwise, it is determined that the noise signal is steady-state noise.

[0074] The specific calculation formula of the noise change rate is as follows:

[0075]

[0076] Among them, represents the noise change rate of the sub-band at time t, and respectively represent the sub-bands at time t and time t - 1 .

[0077] Specifically, the noise type is judged by calculating the noise change rate, effectively distinguishing transient noise and steady-state noise. If the noise change rate is greater than the preset value, it is regarded as transient noise, and the filtering strategy can be flexibly adjusted to enhance the response ability to burst noise, while a more stable processing method is adopted for steady-state noise. In this way, the noise reduction effect can be optimized in different noise environments, improving the clarity and processing efficiency of the audio signal.

[0078] S404: Determine the filter weights of each sub-band according to the error value between the sub-band correction value output by the filter and the actual sub-band value:

[0079]

[0080] Among them, represents the error value of the sub-band at time t, represents the actual sub-band value of the sub-band at time t, and respectively represent the filter weights of the sub-band at time t and t + 1.

[0081] It should be noted that dynamically adjusting the filter weights according to the error between the sub-band correction value and the actual value helps to achieve more accurate noise cancellation. By calculating the error of each sub-band and combining its own adaptive filtering step size, the response of the filter can be optimized to ensure that the filter can be effectively adjusted for different noise situations, thereby improving the noise cancellation effect, minimizing signal distortion, and improving audio quality.

[0082] S405: Synthesize each sub-band obtained by filtering through the filter with an adaptive filtering step size and filter weights to obtain a reverse noise signal.

[0083] It should be noted that by decomposing the noise signal into multiple sub-bands, the noise characteristics of each sub-band can be processed separately, thereby achieving more refined noise suppression. Using the sparse dictionary learning algorithm in the sub-band can effectively extract the sparse representation of the noise. By learning the characteristics of the noise, the noise and the effective signal can be accurately distinguished. In addition, combined with the dynamic adjustment of the adaptive filtering step size and filter weights, the filtering parameters can be flexibly adjusted according to different noise types (transient noise or steady-state noise), so as to ensure more efficient noise cancellation, reduce the distortion and computational overhead of the audio signal, and finally output the optimized audio signal.

[0084] Specifically, the synthesis formula is specifically Among them, represents the full-band reverse noise signal about t obtained after synthesis, and L represents the total number of sub-bands obtained by decomposition.

[0085] S5: Align the noise signal and the reverse noise signal based on the dynamic phase compensation order.

[0086] Among them, the dynamic phase compensation order refers to the number of historical data points used when aligning the noise signal and the reverse noise signal. It determines the number of past samples considered when performing phase alignment on the noise signal and the reverse noise signal. The phase compensation order is usually dynamically adjusted according to the propagation delay of the noise and the characteristics of the audio signal to optimize the signal alignment effect. Based on the dynamic phase compensation order, the system aligns the noise signal and the reverse noise signal. By adjusting the phase compensation order, the system ensures the phase synchronization of the noise and the reverse noise signal, thereby enhancing the noise cancellation effect. Dynamically adjusting the phase compensation order can effectively cope with the changes of different noise signals and optimize the noise reduction process.

[0087] In a possible implementation manner, the calculation formula of the dynamic phase compensation order is specifically:

[0088]

[0089] Among them, represents the propagation delay of the noise signal obtained by testing, represents the wireless audio data acquisition frequency, represents the phase adjustment accuracy related to the wireless audio data frequency, and M represents the dynamic phase compensation order.

[0090] Among them, the phase adjustment accuracy can be set to the reciprocal of the main frequency component in the wireless audio data. The compensation order, which represents the number of historical data points used by the filter. The compensation order determines how many past samples the adaptive filter uses for calculation in order to achieve the best phase alignment. By calculating the dynamic phase compensation order, the precise alignment of the noise signal and the reverse noise signal is ensured. By considering the noise propagation delay, the audio acquisition frequency, and the phase adjustment accuracy, the phase compensation can be dynamically adjusted to optimize the signal synchronization effect. In this way, the noise can be more accurately canceled, and at the same time, the audio distortion caused by phase misalignment can be avoided, improving the noise reduction effect and the sound quality.

[0091] In a possible implementation manner, S5 is specifically: [[ID=]27]

[0092] Align the noise signal and the reverse noise signal through an all-pass filter, and the alignment formula is specifically:

[0093]

[0094] Among them, represents the reverse noise signal at time t after alignment, represents the reverse noise signal when the phase compensation order is m, Denote the phase compensation filtering coefficient under m. Denote taking the phase compensation filtering coefficient when the phase of the noise signal at time t and phase takes the minimum value.

[0095] It should be noted that by aligning the phases of the noise signal and the reverse noise signal through an all-pass filter, it can effectively ensure that the phases of the reverse noise signal and the noise signal are consistent, thereby optimizing the noise cancellation effect. This process dynamically adjusts the phase compensation filtering coefficient to make the phase of the reverse noise signal match the original noise signal, minimizing the noise interference in the audio signal, maintaining the clarity and naturalness of the sound quality, and improving the noise reduction efficiency.

[0096] S6: Synthesize the wireless audio data and the aligned reverse noise signal to obtain optimized wireless audio data.

[0097] It can be understood that the system synthesizes the original wireless audio data and the aligned reverse noise signal. Through this synthesis, the noise components in the original signal are effectively canceled, while the effective part of the original audio is retained, and finally the optimized wireless audio data is obtained. In this way, the optimized signal not only reduces the noise interference but also improves the audio quality.

[0098] In a possible implementation manner, S6 is specifically:

[0099] Synthesize the wireless audio data and the aligned reverse noise signal according to the time stamp to obtain optimized wireless audio data.

[0100] It can be understood that by synthesizing the original wireless audio data and the aligned reverse noise signal according to the time stamp, the noise components are effectively reduced, while the effective information of the audio is retained, thereby improving the audio quality and ensuring the accuracy of noise reduction.

[0101] S7: Output the optimized wireless audio data.

[0102] In the actual application process, first, audio data is collected through a wireless device (such as a microphone or headphones). Then, the autoregressive prediction algorithm is used to determine whether there is noise. If noise is detected, the noise signal is extracted. Next, the reverse noise signal is generated through the sparse dictionary learning algorithm and the adaptive filtering step size to cancel the noise components. Then, the phases of the noise signal and the reverse noise signal are aligned by dynamically adjusting the phase compensation order to ensure the best noise reduction effect. Then, the optimized audio data is synthesized to remove the noise and retain the audio details, and finally the optimized wireless audio data is output. This process effectively improves the audio quality and the noise reduction efficiency and reduces the noise impact.

[0103] The beneficial effects brought by the technical solution provided by the embodiments of the present invention at least include:

[0104] In the embodiments of the present invention, first, the autoregressive prediction algorithm is used to perform pre-noise judgment on the collected wireless audio data. Only when noise is judged to exist can noise reduction be performed. By analyzing the autocorrelation characteristics of the audio data, the presence or absence of noise can be accurately judged, avoiding noise reduction in the absence of noise, thereby improving the noise reduction efficiency and audio quality, effectively avoiding unnecessary noise reduction processing, improving the calculation efficiency, reducing the audio loss caused by global noise reduction, and at the same time improving the accuracy of noise judgment. During the noise reduction process, the sparse dictionary learning algorithm is combined. A reverse noise signal for canceling the noise signal is generated by a filter with an adaptive filtering step size. The sparse dictionary learning algorithm can effectively extract the noise component by learning the noise characteristics and performing sparse representation, and the adaptive filtering step size dynamically adjusts the filter according to the noise change, thereby achieving more accurate noise cancellation and better audio quality. And based on the dynamic phase compensation order, the noise signal and the reverse noise signal are aligned, and then the wireless audio data and the aligned reverse noise signal are synthesized to obtain the optimized wireless audio data and output. It can accurately judge noise, dynamically adjust the filter and extract noise characteristics, thereby effectively improving the accuracy of noise cancellation, avoiding unnecessary noise reduction processing, and at the same time improving the audio quality and processing efficiency.

[0105] Refer to the attached Figure 2 figures, which show the structural schematic diagram of a wireless audio sound effect processing system provided by the present invention.

[0106] The present invention also provides a wireless audio sound effect processing system 20, which is applied to the above-mentioned wireless audio sound effect processing method, and includes:

[0107] A processor 201.

[0108] A memory 202, on which computer-readable instructions are stored. When the computer-readable instructions are executed by the processor 201, the wireless audio sound effect processing method as in the method embodiments is implemented.

[0109] The wireless audio sound effect processing system 20 provided by the present invention can execute the above-mentioned wireless audio sound effect processing method and achieve the same or similar technical effects. To avoid repetition, the present invention will not be elaborated herein.

[0110] The beneficial effects brought by the technical solution provided by the embodiments of the present invention at least include:

[0111] In the embodiments of the present invention, first, an autoregressive prediction algorithm is used to perform pre-noise judgment on the collected wireless audio data. Only when noise is judged to exist can noise reduction be carried out. By analyzing the autocorrelation characteristics of the audio data, the presence or absence of noise can be accurately judged, avoiding noise reduction in the absence of noise, thereby improving the noise reduction efficiency and audio quality, effectively avoiding unnecessary noise reduction processing, enhancing the computing efficiency, reducing the audio loss caused by global noise reduction, and at the same time improving the accuracy of noise judgment. During the process of noise reduction, a sparse dictionary learning algorithm is combined. A reverse noise signal used to cancel the noise signal is generated by a filter with an adaptive filtering step size. The sparse dictionary learning algorithm can effectively extract the noise components by learning the noise characteristics and performing sparse representation, while the adaptive filtering step size dynamically adjusts the filter according to the noise change, thereby achieving more accurate noise cancellation and better audio quality. And the noise signal and the reverse noise signal are aligned based on the dynamic phase compensation order, and then the wireless audio data and the aligned reverse noise signal are synthesized to obtain optimized wireless audio data and output. It can accurately judge noise, dynamically adjust the filter, and extract noise characteristics, thereby effectively improving the accuracy of noise cancellation, avoiding unnecessary noise reduction processing, and at the same time enhancing the audio quality and processing efficiency.

[0112] It should be understood that the processor in the embodiments of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0113] It should also be understood that the memory in the embodiments of the present invention can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchlink dynamic random access memory (SLDRAM), and direct rambus random access memory (DR RAM).

[0114] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0115] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character "" in this document generally indicates that the objects before and after are in an "or" relationship, but it may also represent an "and / or" relationship, which can be understood specifically with reference to the context.

[0116] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following" or its similar expressions refer to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0117] It should be understood that in various embodiments of the present invention, the magnitudes of the sequence numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0118] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0119] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described devices, apparatuses, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0120] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the devices or units can be in an electrical, mechanical, or other form.

[0121] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0122] In addition, the functional units in each embodiment of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0123] When the above-mentioned functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0124] An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the wireless audio sound effect processing method as described in the method embodiment.

[0125] The computer-readable storage medium provided by the present invention can implement the steps and effects of the wireless audio sound effect processing method in the above method embodiment. To avoid repetition, the present invention will not elaborate further.

[0126] The beneficial effects brought by the technical solution provided by the embodiments of the present invention at least include:

[0127] In the embodiment of the present invention, first, the autoregressive prediction algorithm is used to perform pre-noise judgment on the collected wireless audio data. Only when noise is judged to exist can noise reduction be performed. By analyzing the autocorrelation characteristics of the audio data, the existence of noise can be accurately judged, avoiding noise reduction in the case of no noise, thereby improving the noise reduction efficiency and audio quality, effectively avoiding unnecessary noise reduction processing, enhancing the calculation efficiency, reducing the audio loss caused by global noise reduction, and at the same time improving the accuracy of noise judgment. During the noise reduction process, the sparse dictionary learning algorithm is combined. A reverse noise signal used to cancel the noise signal is generated by a filter with an adaptive filtering step size. The sparse dictionary learning algorithm can effectively extract the noise components by learning the noise characteristics and performing sparse representation, while the adaptive filtering step size dynamically adjusts the filter according to the noise change, thereby achieving more accurate noise cancellation and better audio quality. And the noise signal and the reverse noise signal are aligned based on the dynamic phase compensation order, and then the wireless audio data and the aligned reverse noise signal are synthesized to obtain optimized wireless audio data and output. It can accurately judge noise, dynamically adjust the filter, and extract noise characteristics, thereby effectively improving the accuracy of noise cancellation, avoiding unnecessary noise reduction processing, and at the same time enhancing the audio quality and processing efficiency.

[0128] As described above, this is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can hardly think of changes or substitutions, and all of them should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

[0129] The following points need to be explained: (1) The attached drawings of the embodiments of the present invention only relate to the structures involved in the embodiments of the present invention, and other structures can refer to the general design.

[0130] (2) For clarity, in the attached drawings used to describe the embodiments of the present invention, the thickness of the layer or region is enlarged or reduced, that is, these drawings are not drawn according to the actual scale. It can be understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element can be "directly" on or under the other element or there can be an intermediate element.

[0131] (3) Without conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other to obtain new embodiments.

[0132] As above, this is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A wireless audio sound effect processing method, characterized in that the method Including: S1: Collect wireless audio data; S2: Use the autoregressive prediction algorithm to determine whether there is noise in the wireless audio data. If there is noise, proceed to step S3; otherwise, mark the wireless audio data as optimized wireless audio data and proceed to step S7; S3: Extract the noise signal from the wireless audio data; S4: Combine the sparse dictionary learning algorithm and use a filter with an adaptive filtering step size to generate a reverse noise signal for canceling the noise signal; S5: Align the noise signal and the reverse noise signal based on the dynamic phase compensation order; S6: Synthesize the wireless audio data and the aligned reverse noise signal to obtain optimized wireless audio data; S7: Output the optimized wireless audio data; Among them, step S4 specifically includes: S401: Decompose the noise signal into multiple sub-bands; S402: Perform sparse dictionary learning on each sub-band to obtain sparse coefficients corresponding to the noise characteristics; ； Among them, represents the th sub-band obtained by decomposing the noise signal, represents the corresponding dictionary matrix, represents the corresponding sparse coefficient vector, represents the regularization parameter representing the sparsity constraint weight, represents the square of the L2 norm, represents the L1 norm, represents taking the and when the function is minimized; S403: Determine the adaptive filtering step size of the filter according to the sparse coefficient vector; ； Among them, represents the number of non-zero elements in the sparse coefficient vector, Z represents the number of dictionary atoms, and represent the maximum filtering step size and the minimum filtering step size respectively, represents a constant to avoid a zero denominator. In the case where the noise signal is transient noise, the adaptive filtering step size is calculated as , and in the case where the noise signal is stationary noise, the adaptive filtering step size is calculated as ; S404: Determine the filter weights of each sub-band according to the error value between the sub-band correction value output by the filter and the actual value of the sub-band; ； Among them, represents the sub - frequency band the error value at time t, represents the sub - frequency band the actual value of the sub - frequency band at time t, and respectively represent the sub - frequency band the filter weights at time t and time t + 1; S405: Synthesize each sub-band filtered by the filter with the adaptive filtering step size and the filter weights to obtain the reverse noise signal.

2. The wireless audio sound effect processing method according to claim 1, wherein The method of using the autoregressive prediction algorithm in step S2 to determine whether there is noise in the wireless audio data specifically includes: S201: Frame the wireless audio data in combination with the autoregressive order related to the maximum noise length; ； Among them, N w represents the frame length, p represents the autoregressive order of the autoregressive prediction algorithm, and N max represents the maximum noise length, and N h represents the degree of overlap between frames; S202: Calculate the autocorrelation coefficients of each frame of wireless audio data obtained by framing at different lag orders and the noise variance related to the prediction error of the autoregressive prediction algorithm; ； Among them, represents the estimated value of the autocorrelation function of the wireless audio data at each frame under the lag coefficient τ, N represents the total number of samples in the wireless audio data of the current frame, x k and x k-τ respectively represent the k-th wireless audio data value and the (k - τ)-th wireless audio data value in the wireless audio data of the current frame, a i represents the autocorrelation coefficient of the wireless audio data of the current frame at the i-th order, i = 1, 2, p, represents the noise variance, e k represents the prediction error of the k-th wireless audio data value in the wireless audio data of the current frame; S203: Combine the calculated autocorrelation coefficients to calculate the detection signal difference between the actual value of the wireless audio data and the predicted value of the wireless audio data; ； Among them, represents the actual value of the wireless audio data at the current frame at time t, represents the value of the wireless audio data with a lag of i, represents the difference in the detection signal at time t; S204: Use the noise standard deviation to determine whether there is noise in the wireless audio data according to the detection signal difference; ； Among them, I represents an exponential variable. When I = 1, it means that there is noise in the wireless audio data. When I = 0, it means that there is no noise in the wireless audio data. represents the noise standard deviation, represents the noise screening threshold, .

3. The wireless audio sound effect processing method according to claim 2, wherein Step S3 is specifically: Extract the detection signal difference as the noise signal.

4. The wireless audio sound effect processing method according to claim 3, wherein In the case where the noise change rate is greater than the preset noise change rate, determine the noise signal as transient noise; otherwise, determine the noise signal as steady-state noise; The specific calculation formula of the noise change rate is: ； Among them, represents the noise change rate of the sub-band at time t, and respectively represent the sub-band l at time t and time t-1.

5. The wireless audio sound effect processing method according to claim 1, wherein The specific calculation formula of the dynamic phase compensation order is: ； wherein, represents the propagation delay of the measured noise signal, represents the wireless audio data acquisition frequency, represents the phase adjustment accuracy related to the wireless audio data frequency, and M represents the dynamic phase compensation order.

6. The wireless audio sound effect processing method according to claim 5, wherein Step S5 is specifically: Align the noise signal and the reverse noise signal through an all-pass filter, and the alignment formula is specifically: ； Among them, represents the reverse noise signal at time t after alignment, represents the reverse noise signal when the phase compensation order is m, represents the phase compensation filter coefficient under m, represents taking the phase compensation filter coefficient when the phase of the noise signal at time t and phase takes the minimum value.

7. The wireless audio sound effect processing method according to claim 1, wherein Step S6 is specifically: Synthesize the wireless audio data and the aligned reverse noise signal according to the timestamp to obtain the optimized wireless audio data.

8. A wireless audio sound effect processing system, characterized in that, Including: A processor; A memory, on which computer-readable instructions are stored. When the computer-readable instructions are executed by the processor, the wireless audio sound processing method according to any one of claims 1 to 7 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, the wireless audio sound processing method according to any one of claims 1 to 7 is implemented.

Images