WO2024093307A1 - 音频处理方法、装置、电子设备和计算机可读存储介质 - Google Patents

音频处理方法、装置、电子设备和计算机可读存储介质 Download PDF

Info

Publication number
WO2024093307A1
WO2024093307A1 PCT/CN2023/103752 CN2023103752W WO2024093307A1 WO 2024093307 A1 WO2024093307 A1 WO 2024093307A1 CN 2023103752 W CN2023103752 W CN 2023103752W WO 2024093307 A1 WO2024093307 A1 WO 2024093307A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
audio
audio output
sub
modulation
Prior art date
Application number
PCT/CN2023/103752
Other languages
English (en)
French (fr)
Inventor
李倩
Original Assignee
恒玄科技(上海)股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 恒玄科技(上海)股份有限公司 filed Critical 恒玄科技(上海)股份有限公司
Publication of WO2024093307A1 publication Critical patent/WO2024093307A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R3/00—Circuits for transducers, loudspeakers or microphones
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R3/00—Circuits for transducers, loudspeakers or microphones
    • H04R3/02—Circuits for transducers, loudspeakers or microphones for preventing acoustic reaction, i.e. acoustic oscillatory feedback

Definitions

  • the present invention relates to the field of voice communication, and in particular to an audio processing method, device, electronic device and computer-readable storage medium.
  • audio devices such as headphones, loudspeakers, and assistive listening products have been widely used.
  • microphones and speakers are installed at the same time in existing audio devices.
  • the sound played by the speaker will be picked up by the microphone, and the sound path forms a closed loop.
  • the signal is continuously superimposed and amplified in the sound feedback loop to form positive feedback, resulting in a single-frequency howling.
  • the existing echo suppression scheme needs to eliminate the feedback signal from the speaker to the microphone from the audio signal collected by the microphone.
  • the feedback path is short, and the correlation between the microphone input signal and the reference signal played by the speaker is strong.
  • the existing echo suppression scheme will cause the wearer's own voice signal played by the speaker to be distorted, affecting the user experience.
  • Objectives of the present invention include, for example, providing an audio processing method, apparatus, electronic device, and computer-readable storage medium, which can reduce the distortion of an audio signal played by a speaker while eliminating howling.
  • the present invention provides an audio processing method, which is applied to an audio processing device including an audio input device and an audio output device, the method comprising: obtaining an audio input signal collected by the audio input device in a current sampling period, performing audio signal processing on the audio input signal to obtain an audio output signal, and playing the audio output signal via the audio output device; the audio signal processing at least comprises: obtaining a previous audio output signal, the previous audio output signal being the audio output signal in a previous sampling period, A feedback signal is obtained according to the front audio output signal, signal compensation is performed on the audio input signal according to the feedback signal to obtain a compensated audio signal, and a phase of the compensated audio signal is nonlinearly modulated.
  • the audio signal processing further includes: acquiring a front-compensated audio signal, the front-compensated audio signal being a compensated audio signal in a previous sampling period; calculating feedback frequency response data according to the front audio output signal and the front-compensated audio signal, and determining whether the feedback frequency response data is greater than preset frequency response data; if the feedback frequency response data is greater than the preset frequency response data, performing nonlinear modulation on the phase of the compensated audio signal.
  • the calculating the feedback frequency response data according to the front audio output signal and the front compensated audio signal includes: acquiring the front feedback frequency response data, where the front feedback frequency response data is the feedback frequency response data of a previous sampling period; and calculating the feedback frequency response data of a current sampling period according to the front audio output signal, the compensated audio signal, and the front feedback frequency response data.
  • the calculating the feedback frequency response data of the current sampling period according to the front audio output signal, the front compensated audio signal, and the front feedback frequency response data includes: determining a sampling number K according to a preset feedback path length; acquiring the audio output signals of K sampling periods adjacent to the current sampling period to form an audio output array; and calculating the feedback frequency response data of the current sampling period according to the audio output array, the compensated audio signal, and the front feedback frequency response data.
  • the nonlinearly modulating the phase of the compensated audio signal includes: decomposing the compensated audio signal into sub-bands according to the frequency to obtain a plurality of sub-band signals; nonlinearly modulating the phase of each of the sub-band signals according to a first modulation method to obtain a plurality of modulated sub-band signals; and synthesizing the plurality of modulated sub-band signals.
  • the nonlinearly modulating the phase of the compensated audio signal includes: performing sub-band decomposition on the compensated audio signal according to frequency to obtain multiple sub-band signals; performing convolution and limiting processing on each of the sub-band signals respectively; performing signal synthesis on the multiple sub-band signals after the convolution and limiting processing to obtain a synthesized signal; and performing nonlinear modulation on the phase of the synthesized signal according to a first modulation method.
  • the first modulation method includes: obtaining a first modulation parameter and a second modulation parameter, constructing a first modulation formula based on the first modulation parameter and the second modulation parameter, the first modulation parameter being used to adjust the degree of nonlinearity of the first modulation formula, and the second modulation parameter being used to adjust the slope of the first modulation formula; and using the first modulation formula to perform nonlinear modulation.
  • the nonlinearly modulating the phase of the compensated audio signal includes: performing sub-band decomposition on the compensated audio signal according to frequency to obtain multiple sub-band signals; obtaining the sub-band signal having a frequency greater than a preset frequency as a target sub-band signal; performing nonlinear modulation on the phase of each of the target sub-band signals according to a second modulation method to obtain multiple modulated sub-band signals; and performing signal synthesis on the sub-band signals other than the target sub-band signal and the modulated sub-band signal.
  • the nonlinearly modulating the phase of the compensated audio signal includes: performing sub-band decomposition on the compensated audio signal according to frequency to obtain multiple sub-band signals; obtaining the sub-band signal with a frequency greater than a preset frequency as a target sub-band signal; synthesizing the target sub-band signal to obtain a first synthesized signal; nonlinearly modulating the phase of the first synthesized signal according to a second modulation method to obtain a first modulated synthesized signal; and synthesizing the sub-band signals except the target sub-band signal and the first modulated synthesized signal.
  • the second modulation method includes: obtaining a first modulation parameter, constructing a second modulation formula based on the first modulation parameter, wherein the first modulation parameter is used to adjust the degree of nonlinearity of the second modulation formula; and performing nonlinear modulation using the second modulation formula.
  • the present invention provides an audio processing device, comprising: an audio input module, the audio input module is used to collect audio input signals; a processing module, the processing module is used to obtain the audio input signals collected by the audio input module in the current sampling period, and perform audio signal processing on the audio input signals to obtain audio output signals, wherein the audio signal processing at least includes: obtaining a front audio output signal, the front audio output signal is the audio output signal in the previous sampling period, obtaining a feedback signal according to the front audio output signal, performing signal compensation on the audio input signal according to the feedback signal to obtain a compensated audio signal, and performing nonlinear modulation on the phase of the compensated audio signal; an audio output module, the audio output module is used to play the audio output signal; and a storage module is used to store the front audio output signal. Number.
  • the present invention provides an electronic device comprising: an audio input device, an audio output device, at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the audio processing method as described in any one of the aforementioned embodiments.
  • the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the audio processing method described in any one of the aforementioned embodiments.
  • the audio processing method, device, electronic device and computer-readable storage medium provided by the embodiments of the present invention, after the audio input device collects the audio input signal, a feedback signal is obtained according to the previous audio output signal, the previous audio output signal is the audio output signal played via the audio output device in the previous sampling period, the feedback signal obtained according to the previous audio output signal is the signal formed by the part of the estimated previous audio output signal that is played via the audio output device and then re-collected by the audio input device, and the audio input signal is compensated according to the feedback signal to eliminate the part of the audio played by the audio output device contained in the audio input signal that is re-collected, the compensated audio signal obtained in this way is the audio signal of the actual ambient audio, and the compensated audio signal is used as the audio Playing the audio output signal can effectively eliminate the possible howling; in addition, since audio processing is a continuous process, the audio output signal obtained in the current sampling period will be used as the previous audio output signal at the next sampling moment to perform audio signal processing on
  • FIG1 is a schematic diagram of a flow chart of an audio processing method provided in Embodiment 1;
  • FIG2 is a schematic diagram of a flow chart of audio signal processing in the audio processing method provided in Embodiment 1;
  • FIG3 is a schematic diagram of a frequency-phase response image of a first modulation formula in the audio processing method provided in Embodiment 1;
  • FIG4 is a schematic diagram of a frequency-phase response image of a second modulation formula in the audio processing method provided in Embodiment 1;
  • FIG5 is a schematic diagram of a flow chart of audio signal processing in an audio processing method provided by another embodiment
  • FIG6 is a schematic diagram of the structure of an audio processing device provided in Embodiment 2;
  • FIG. 7 is a schematic diagram of the structure of an electronic device provided in the third embodiment.
  • Embodiment 1 of the present invention provides an audio processing method, which is applied to an audio processing device including an audio input device and an audio output device, as shown in FIG1 , and includes the following steps:
  • Step S101 obtaining an audio input signal collected by an audio input device during a current sampling period.
  • the audio input device continuously collects external audio, and integrates the external audio data collected within a preset sampling period to form an audio input signal.
  • Step S102 performing audio signal processing on the audio input signal to obtain an audio output signal.
  • the audio signal processing process is shown in FIG2 and includes the following steps:
  • Step S201 Acquire the front audio output signal.
  • the previous audio output signal is the audio output signal of the sampling period before the current sampling period.
  • Step S202 obtaining a feedback signal according to the front audio output signal.
  • the algorithm filter can be written using the normalized LMS algorithm to establish a filter prediction feedback path, and the front audio output signal is convolved with the filter prediction feedback path to obtain a feedback signal.
  • the feedback signal is the audio signal that may be re-collected by the audio input device after the estimated front audio output signal is played by the audio output device. It can be understood that the above is only a specific example of obtaining the feedback signal in this embodiment, and does not constitute a limitation.
  • other methods may also be used, such as pre-measuring the audio feedback parameters between the audio input device and the audio output device by a measuring device, and obtaining a feedback signal according to the audio feedback parameters that have been measured during use, and other methods can be used flexibly according to actual needs.
  • Step S203 performing signal compensation on the audio input signal according to the feedback signal to obtain a compensated audio signal.
  • the feedback signal is removed from the audio input signal at the current sampling moment, so as to obtain a compensated audio signal after the echo howling is eliminated.
  • Step S204 nonlinearly modulate the phase of the compensated audio signal.
  • the compensated audio signal is pre-decomposed into sub-bands according to the frequency size, and a plurality of sub-band signals of different frequencies are obtained after the decomposition, and then the phase of each sub-band signal is non-linearly modulated according to the first modulation method to obtain a plurality of modulated sub-band signals; the plurality of modulated sub-band signals are synthesized to obtain the audio output signal of the current sampling period.
  • the first modulation method is to use the first modulation formula to non-linearly modulate the phase of each sub-band signal, and the first modulation formula includes a first modulation parameter and a second modulation parameter, the first modulation parameter is used to adjust the non-linear degree of the first modulation formula, and the second modulation parameter is used to adjust the slope of the first modulation formula.
  • the first modulation formula is:
  • ⁇ is the first modulation parameter
  • the size of ⁇ affects the nonlinearity of the first modulation formula
  • ⁇ is the second modulation parameter
  • the size of ⁇ affects the slope of the first modulation formula
  • H(z) is the phase size to be adjusted corresponding to different frequencies
  • Z is the frequency
  • N is a preset constant.
  • FIG3 it is a schematic diagram of the frequency-phase response image of the first modulation formula when ⁇ is set to 0.4 and ⁇ is set to 0.6.
  • the nonlinear degree of phase adjustment of the first modulation formula in the high-frequency part (the part greater than 1.0 ⁇ 10 4 Hz) is greater than that in the low-frequency part (the part less than 0.5 ⁇ 10 4 Hz). Since the human auditory system is usually more sensitive to nonlinear phase changes at low frequencies, but not to nonlinear phase changes at high frequencies. Therefore, the phase adjustment amplitude of the first modulation formula in the high-frequency part is greater than the phase adjustment amplitude in the low-frequency part. Larger, it can achieve a better effect of reducing the correlation between the audio input signal and the audio output signal without changing the listening experience.
  • the compensated audio signal may be decomposed into sub-bands according to the frequency to obtain multiple sub-band signals; convolution and limiting processing may be performed on each sub-band signal respectively; the multiple sub-band signals after convolution and limiting processing may be synthesized to obtain a synthesized signal; and the phase of the synthesized signal may be nonlinearly modulated according to the first modulation method.
  • the multiple sub-band signals after convolution and limiting processing may be synthesized to obtain a synthesized signal, and then the phase of the synthesized signal may be nonlinearly modulated using the first modulation method, which can reduce the amount of calculation in the phase modulation process and improve the overall audio processing efficiency.
  • the compensated audio signal may be decomposed into sub-bands according to the frequency to obtain multiple sub-band signals; a sub-band signal with a frequency greater than a preset frequency is obtained as a target sub-band signal; the phase of each target sub-band signal is nonlinearly modulated according to a second modulation method to obtain multiple modulated sub-band signals; and the sub-band signals other than the target sub-band signal and the modulated sub-band signal are synthesized.
  • the second modulation method is to use a second modulation formula to nonlinearly modulate the phase of each target sub-band signal, and the second modulation formula includes a first modulation parameter, and the first modulation parameter is used to adjust the nonlinearity of the first modulation formula.
  • the second modulation formula is:
  • ⁇ is the first modulation parameter
  • the size of ⁇ affects the nonlinearity of the first modulation formula
  • H(z) is the phase size to be adjusted corresponding to different frequencies
  • Z is the frequency
  • N is a preset constant.
  • FIG. 4 a schematic diagram of the frequency-phase response image of the second modulation formula when ⁇ is set to 0.4 is shown.
  • the compensated audio signal may be decomposed into sub-bands according to the frequency to obtain a plurality of sub-band signals; the sub-band signals having a frequency greater than a preset frequency are synthesized to obtain a first synthesized signal; and the phase of the first synthesized signal is modulated according to the second modulation method.
  • the first synthesized signal is obtained by performing nonlinear modulation on the subband signal except the target subband signal and the first synthesized signal.
  • the subband signal with a frequency greater than the preset frequency is synthesized to obtain the first synthesized signal; the phase of the first synthesized signal is nonlinearly modulated according to the second modulation method, which can reduce the amount of calculation in the phase modulation process and improve the overall audio processing efficiency.
  • the aforementioned steps S201 to S204 are merely examples of the audio signal processing process in one embodiment of the present invention.
  • the audio signal processing process may also be as shown in FIG. 5 , including the following steps.
  • Step S301 Acquire the front audio output signal.
  • Step S302 obtaining a feedback signal according to the front audio output signal.
  • Step S303 performing signal compensation on the audio input signal according to the feedback signal to obtain a compensated audio signal.
  • steps S301 to S303 in this embodiment are substantially the same as steps S201 to S203 in the previous embodiment, and specific details may be referred to the specific description of the previous embodiment.
  • Step S304 Acquire a pre-compensated audio signal.
  • the previous compensated audio signal is a compensated audio signal of a previous sampling period.
  • Step S305 Calculate and obtain feedback frequency response data according to the front audio output signal and the front compensated audio signal.
  • front feedback frequency response data can be obtained, where the front feedback frequency response data is the feedback frequency response data of the previous sampling period; the feedback frequency response data of the current sampling period is calculated based on the front audio output signal, the front compensated audio signal, and the front feedback frequency response data.
  • W(n) [w 0 (n), w 1 (n), w 2 (n), ..., w L-1 (n)] T
  • w(n) is the estimated feedback path
  • n is the sampling period
  • W(n) is the feedback frequency of the previous sampling period
  • r(n) is the compensated audio signal of the previous sampling period
  • r(n-L+1) is the compensated audio signal of the previous L+2 sampling periods
  • L is the length of the feedback path from the audio output of the preset audio output device to the audio input device.
  • Step S306 Determine whether the feedback frequency response data is greater than the preset frequency response data, if so, execute step S307, if not, execute step S308.
  • the preset frequency response data is the threshold data for generating the howling risk. If the feedback frequency response data is greater than the preset frequency response data, it indicates that the howling risk may occur, and step S306 is executed at this time. On the contrary, if the feedback frequency response data is not greater than the preset frequency response data, it indicates that the howling risk will not occur, and step S307 is executed at this time.
  • Step S307 nonlinearly modulate the phase of the compensated audio signal.
  • Step S308 using the compensated audio signal as an audio output signal.
  • the feedback frequency response data is calculated before the phase of the compensation audio signal is nonlinearly modulated.
  • the feedback frequency response data is greater than the preset frequency response data, indicating that a howling risk may occur
  • the phase of the compensation audio signal is nonlinearly modulated. Otherwise, the phase of the compensation audio signal is not nonlinearly modulated.
  • the number of times the phase of the compensation audio signal is modulated is reduced, the amount of phase modulation calculation is reduced, and the efficiency of audio signal processing is improved while ensuring that no howling occurs.
  • Step S103 Play the audio output signal via the audio output device.
  • the audio processing method provided in the first embodiment of the present invention, after the audio input device collects the audio input signal, a feedback signal is obtained according to the previous audio output signal, the previous audio output signal is the audio output signal played by the audio output device in the previous sampling period, and the feedback signal obtained according to the previous audio output signal is the estimated previous audio output signal played by the audio output device in the previous sampling period.
  • the signal formed by the part re-collected by the audio input device after the audio output device plays can be compensated for the audio input signal according to the feedback signal, so that the part of the audio played by the audio output device contained in the audio input signal that is re-collected can be eliminated.
  • the compensated audio signal obtained in this way is the audio signal of the actual ambient audio.
  • Playing the compensated audio signal as the audio output signal can effectively eliminate the possible howling.
  • the audio output signal obtained in the current sampling period will be used as the previous audio output signal at the next sampling moment to perform audio signal processing on the audio input signal collected at the next sampling moment, and the phase of the compensated audio signal is nonlinearly modulated, which can increase the difference between the compensated audio signal and the audio input signal collected at the next sampling moment, thereby avoiding the audio content similarity between the compensated audio signal and the audio input signal collected at the next sampling moment being too high, resulting in excessive audio content being eliminated when the audio input signal collected at the next sampling moment is compensated, thereby reducing the audio distortion that may be caused by the signal compensation process, and achieving the effect of reducing audio signal distortion while eliminating howling.
  • Embodiment 2 of the present invention provides an audio processing device, as shown in FIG6 , including: an audio input module 601, the audio input module 601 is used to collect audio input signals; a processing module 602, the processing module 602 is used to obtain the audio input signal collected by the audio input module 601 in the current sampling period, and perform audio signal processing on the audio input signal to obtain an audio output signal, wherein the audio signal processing at least includes: obtaining a previous audio output signal, the previous audio output signal is the audio output signal of the previous sampling period, obtaining a feedback signal according to the previous audio output signal, performing signal compensation on the audio input signal according to the feedback signal to obtain a compensated audio signal, and performing nonlinear modulation on the phase of the compensated audio signal; an audio output module 603, the audio output module 603 is used to play the audio output signal; and a storage module 604, the storage module 604 is used to store the previous audio output signal.
  • the processing module 602 obtains the feedback signal according to the previous audio output signal stored in the storage module 604, the previous audio output signal is the audio output signal played via the audio output module 603 in the previous sampling period stored in the storage module 604, the feedback signal obtained according to the previous audio output signal is the signal formed by the part of the estimated previous audio output signal re-collected by the audio input module 601 after being played via the audio output module 603, and the audio input signal is compensated according to the feedback signal to eliminate the audio input signal.
  • the audio output module 603 includes a re-collected portion of the audio played, and the compensated audio signal obtained in this way is the audio signal of the actual ambient audio.
  • the processing module 602 plays the compensated audio signal as an audio output signal, which can effectively eliminate the possible howling.
  • the audio output signal obtained in the current sampling period will be used as the previous audio output signal at the next sampling moment to perform audio signal processing on the audio input signal collected at the next sampling moment, and the phase of the compensated audio signal is nonlinearly modulated, which can increase the difference between the compensated audio signal and the audio input signal collected at the next sampling moment, thereby avoiding the audio content similarity between the compensated audio signal and the audio input signal collected at the next sampling moment being too high, resulting in excessive audio content being eliminated when the audio input signal collected at the next sampling moment is compensated, thereby reducing the audio distortion that may be caused by the signal compensation process, and achieving the effect of reducing audio signal distortion while eliminating howling.
  • Embodiment 3 of the present invention relates to an electronic device, as shown in FIG7 , comprising: an audio input device 701, an audio output device 702, at least one processor 703; and a memory 704 communicatively connected to the at least one processor 703; wherein the memory 704 stores instructions executable by the at least one processor 703, and the instructions are executed by the at least one processor 703 so that the at least one processor 703 can execute the audio processing methods in the above-mentioned embodiments.
  • the memory and the processor are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors and memories together.
  • the bus can also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are all well known in the art and are therefore not further described herein.
  • the bus interface provides an interface between the bus and the transceiver.
  • the transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium.
  • the data processed by the processor is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor.
  • the processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions.
  • Memory can be used to store data used by the processor when performing operations.
  • Embodiment 4 of the present invention relates to a computer-readable storage medium storing a computer program.
  • the computer program is executed by a processor, the above method embodiment is implemented.
  • a storage medium including a number of instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to execute all or part of the steps of each embodiment method of the present application.
  • the aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Otolaryngology (AREA)
  • Circuit For Audible Band Transducer (AREA)

Abstract

本申请提供了一种音频处理方法、装置、电子设备及计算机可读存储介质，涉及语音处理领域。其中，音频处理方法包括：获取当前采样时段的音频输入信号，对音频输入信号进行音频信号处理得到音频输出信号，将音频输出信号进行播放；音频信号处理包括：获取前音频输出信号，前音频输出信号为前一采样时段的音频输出信号，根据前音频输出信号获取反馈信号，根据反馈信号对音频输入信号进行信号补偿得到补偿音频信号，对补偿音频信号的相位进行非线性调制。与现有技术相比，本申请所提供的音频处理方法、装置、电子设备和计算机可读存储介质，能够在消除啸叫的同时减少扬声器播放的音频信号失真。

Description

音频处理方法、装置、电子设备和计算机可读存储介质 技术领域
本发明涉及语音通信领域，具体而言，涉及一种音频处理方法、装置、电子设备和计算机可读存储介质。
背景技术
随着技术的发展，耳机、扩音器、辅听产品等音频设备得到广泛的应用。然而，为了达到降噪、扩音、辅听等不同功能，在现有的音频设备中会同时安装配备麦克风和扬声器。在这种同时安装配备麦克风和扬声器的音频设备中，由扬声器播放的声音会被麦克风回采，声通路形成闭合回路，信号在声反馈回路中不断叠加放大形成正反馈，产生单频的啸叫。为了从根本上消除啸叫，现有的回声抑制方案需要将扬声器反馈到麦克风的反馈信号从麦克风采集的音频信号中消除，但是由于麦克风和扬声器之间距离较近，反馈路径较短，麦克风输入信号和扬声器播放的参考信号的相关性较强，现有的回声抑制方案会导致扬声器播放的佩戴者自己的语音信号失真，影响用户体验。
发明内容
本发明的目的包括，例如，提供了一种音频处理方法、装置、电子设备和计算机可读存储介质，其能够在消除啸叫的同时减少扬声器播放的音频信号失真。
本发明的实施例可以这样实现：第一方面，本发明提供一种音频处理方法，应用于包括音频输入设备和音频输出设备的音频处理装置，所述方法包括：获取当前采样时段所述音频输入设备采集的音频输入信号，对所述音频输入信号进行音频信号处理得到音频输出信号，将所述音频输出信号经由所述音频输出设备进行播放；所述音频信号处理至少包括：获取前音频输出信号，所述前音频输出信号为前一采样时段的所述音频输出信号， 根据所述前音频输出信号获取反馈信号，根据所述反馈信号对所述音频输入信号进行信号补偿得到补偿音频信号，对所述补偿音频信号的相位进行非线性调制。
在可选的实施方式中，所述音频信号处理还包括：获取前补偿音频信号，所述前补偿音频信号为前一采样时段的补偿音频信号；根据所述前音频输出信号和所述前补偿音频信号计算得到反馈频响数据，判断所述反馈频响数据是否大于预设频响数据；若所述反馈频响数据大于预设频响数据，执行对所述补偿音频信号的相位进行非线性调制。
在可选的实施方式中，所述根据所述前音频输出信号和所述前补偿音频信号计算得到反馈频响数据，包括：获取前反馈频响数据，所述前反馈频响数据为前一采样时段的所述反馈频响数据；根据所述前音频输出信号、所述补偿音频信号、以及所述前反馈频响数据计算得到当前采样时段的所述反馈频响数据。
在可选的实施方式中，所述根据所述前音频输出信号、所述前补偿音频信号、以及所述前反馈频响数据计算得到当前采样时段的所述反馈频响数据，包括：根据预设反馈路径长度确定采样数量K；获取与当前采样时段相邻的K个采样时段的所述音频输出信号形成音频输出阵列；根据所述音频输出阵列、所述补偿音频信号、以及所述前反馈频响数据计算得到当前采样时段的所述反馈频响数据。
在可选的实施方式中，所述对所述补偿音频信号的相位进行非线性调制，包括：根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；根据第一调制方法分别对各个所述子带信号的相位进行非线性调制，得到多个调制子带信号；将所述多个调制子带信号进行信号合成。
在可选的实施方式中，所述对所述补偿音频信号的相位进行非线性调制，包括：根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；分别对各个所述子带信号进行卷积和限幅处理；将所述卷积和限幅处理后的所述多个子带信号进行信号合成，得到合成信号；根据第一调制方法对所述合成信号的相位进行非线性调制。
在可选的实施方式中，所述第一调制方法，包括：获取第一调制参数和第二调制参数，根据所述第一调制参数和所述第二调制参数构建第一调制公式，所述第一调制参数用于调整所述第一调制公式的非线性程度，所述第二调制参数用于调整所述第一调制公式的斜率；使用所述第一调制公式进行非线性调制。
在可选的实施方式中，所述对所述补偿音频信号的相位进行非线性调制，包括：根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；获取频率大于预设频率的所述子带信号作为目标子带信号；根据第二调制方法分别对各个所述目标子带信号的相位进行非线性调制，得到多个调制子带信号；将除所述目标子带信号外的所述子带信号和所述调制子带信号进行信号合成。
在可选的实施方式中，所述对所述补偿音频信号的相位进行非线性调制，包括：根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；获取频率大于预设频率的所述子带信号作为目标子带信号；将所述目标子带信号进行信号合成，得到第一合成信号；根据第二调制方法对第一合成信号的相位进行非线性调制，得到第一调制合成信号；将除所述目标子带信号外的所述子带信号和所述第一调制合成信号进行信号合成。
在可选的实施方式中，所述第二调制方法，包括：获取第一调制参数，根据所述第一调制参数构建第二调制公式，所述第一调制参数用于调整所述第二调制公式的非线性程度；使用所述第二调制公式进行非线性调制。
第二方面，本发明提供一种音频处理装置，包括：音频输入模块，所述音频输入模块用于采集音频输入信号；处理模块，所述处理模块用于获取当前采样时段所述音频输入模块采集的音频输入信号，对所述音频输入信号进行音频信号处理得到音频输出信号，所述音频信号处理至少包括：获取前音频输出信号，所述前音频输出信号为前一采样时段的所述音频输出信号，根据所述前音频输出信号获取反馈信号，根据所述反馈信号对所述音频输入信号进行信号补偿得到补偿音频信号，对所述补偿音频信号的相位进行非线性调制；音频输出模块，所述音频输出模块用于对所述音频输出信号进行播放；存储模块，所述存储模块用于存储所述前音频输出信 号。
第三方面，本发明提供一种电子设备，包括：音频输入设备、音频输出设备、至少一个处理器；以及，与所述至少一个处理器通信连接的存储器；其中，所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行如前述实施方式中任意一项所述的音频处理方法。
第四方面，本发明提供一种计算机可读存储介质，存储有计算机程序，所述计算机程序被处理器执行实现前述实施方式中任意一项所述的音频处理方法。
与现有技术相比，本发明实施例所提供的音频处理方法、装置、电子设备和计算机可读存储介质中，音频输入设备采集到音频输入信号后，根据前音频输出信号获取反馈信号，前音频输出信号即为前一采样时段经由音频输出设备进行播放的音频输出信号，根据前音频输出信号获取得到的反馈信号即为预估的前音频输出信号经由音频输出设备进行播放后被音频输入设备重新采集的部分形成的信号，根据反馈信号对音频输入信号进行信号补偿，即可消除音频输入信号中所包含的音频输出设备播放的音频被重新采集的部分，如此得到的补偿音频信号即为实际的环境音频的音频信号，将补偿音频信号作为音频输出信号进行播放既可有效的消除可能产生的啸叫；此外，由于音频处理是一个连续的过程，当前采样时段得到的音频输出信号会在下一采样时刻作为前音频输出信号对下一采样时刻采集到的音频输入信号进行音频信号处理，对补偿音频信号的相位进行非线性调制，可以增大补偿音频信号与下一采样时刻采集到的音频输入信号的差异度，从而避免补偿音频信号与下一采样时刻采集到的音频输入信号的音频内容相似度过高、导致对下一采样时刻采集到的音频输入信号进行信号补偿时消除了过多的音频内容，从而减少信号补偿过程可能导致的音频失真，实现在消除啸叫的同时减少音频信号失真的效果。
附图说明
为了更清楚地说明本发明实施例的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，应当理解，以下附图仅示出了本发明的某些实施例，因此不应被看作是对范围的限定，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他相关的附图。
图1为实施例一所提供的音频处理方法的流程示意图；
图2为实施例一所提供的音频处理方法中音频信号处理的流程示意图；
图3为实施例一所提供的音频处理方法中第一调制公式的频率-相位响应图像的示意图；
图4为实施例一所提供的音频处理方法中第二调制公式的频率-相位响应图像的示意图；
图5为另一实施例所提供的音频处理方法中音频信号处理的流程示意图；
图6为实施例二所提供的音频处理装置的结构示意图；
图7为实施例三所提供的电子设备的结构示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。通常在此处附图中描述和示出的本发明实施例的组件可以以各种不同的配置来布置和设计。
因此，以下对在附图中提供的本发明的实施例的详细描述并非旨在限制要求保护的本发明的范围，而是仅仅表示本发明的选定实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
应注意到：相似的标号和字母在下面的附图中表示类似项，因此，一旦某一项在一个附图中被定义，则在随后的附图中不需要对其进行进一步 定义和解释。
在本发明的描述中，需要说明的是，若出现术语“上”、“下”、“内”、“外”等指示的方位或位置关系为基于附图所示的方位或位置关系，或者是该发明产品使用时惯常摆放的方位或位置关系，仅是为了便于描述本发明和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本发明的限制。
此外，若出现术语“第一”、“第二”等仅用于区分描述，而不能理解为指示或暗示相对重要性。
需要说明的是，在不冲突的情况下，本发明的实施例中的特征可以相互结合。
本发明实施例一提供了一种音频处理方法，应用于包括音频输入设备和音频输出设备的音频处理装置，如图1所示，包括以下步骤：
步骤S101：获取当前采样时段音频输入设备采集的音频输入信号。
在本步骤中，音频输入设备持续采集外界音频，每间隔一个预设的采样时长将这一采样时段内采集的外界音频数据整合形成音频输入信号。
步骤S102：对音频输入信号进行音频信号处理得到音频输出信号。
在本步骤中，音频信号处理过程如图2所示，包括以下步骤：
步骤S201：获取前音频输出信号。
具体的，前音频输出信号即为当前采样时段的前一采样时段的音频输出信号。
步骤S202：根据前音频输出信号获取反馈信号。
具体的，在本实施例中，可以用归一化LMS算法编写算法滤波器，从而建立滤波预估反馈路径，将前音频输出信号与滤波预估反馈路径相卷积，得到反馈信号。反馈信号即为预估的前音频输出信号经由音频输出设备播放后可能被音频输入设备重新采集的音频信号。可以理解的是，前述仅为本实施例中获取反馈信号的一种具体的举例说明，并不构成限定，在本发 明的其它实施例中，也可以使用其它方法，例如通过测量设备预先测量音频输入设备和音频输出设备之间的音频反馈参数，使用时根据已经测量得到的音频反馈参数得到反馈信号等其它方法，具体可以根据实际需要进行灵活的使用。
步骤S203：根据反馈信号对音频输入信号进行信号补偿得到补偿音频信号。
具体的，在本步骤中，将反馈信号从当前采样时刻的音频输入信号中去除，即可得到回声啸叫消除后的补偿音频信号。
步骤S204：对补偿音频信号的相位进行非线性调制。
具体的，在本步骤中，预先根据频率大小对补偿音频信号进行子带分解，分解后得到多个不同频率的子带信号，然后根据第一调制方法分别对各个子带信号的相位进行非线性调制，得到多个调制子带信号；将多个调制子带信号进行信号合成得到当前采样时段的音频输出信号。其中，第一调制方法为使用第一调制公式对各个子带信号的相位进行非线性调制，第一调制公式包括第一调制参数和第二调制参数，第一调制参数用于调整第一调制公式的非线性程度，第二调制参数用于调整第一调制公式的斜率。
进一步的，在本发明的一种实施例中，第一调制公式为：
其中，α为第一调制参数，α的大小影响第一调制公式的非线性程度，β为第二调制参数，β的大小影响第一调制公式的斜率，H(z)为不同频率对应的需调整的相位大小，Z为频率，N为预设常数。
具体的，如图3所示为α设置为0.4、β设置为0.6时，第一调制公式的频率-相位响应图像的示意图。如图3所示，第一调制公式在高频部分(大于1.0×104Hz的部分)的相位调节非线性程度较之低频部分(小于0.5×104Hz的部分)的相位调节非线性程度更大，由于人类听觉系统通常对于低频的相位非线性变化比较敏感，而对于高频的相位非线性变化不敏感。因此第一调制公式在高频部分的相位调节幅度较之低频部分的相位调节幅度 更大，可以在不改变听感的情况下达到一个比较好的减弱音频输入信号和音频输出信号之间相关性的效果。
在本发明的另一实施例中，也可以是根据频率大小对补偿音频信号进行子带分解，得到多个子带信号；分别对各个子带信号进行卷积和限幅处理；将卷积和限幅处理后的多个子带信号进行信号合成，得到合成信号；根据第一调制方法对合成信号的相位进行非线性调制。将卷积和限幅处理后的多个子带信号进行信号合成，得到合成信号，然后使用第一调制方法对合成信号的相位进行非线性调制，可以减少相位调制过程中的计算量，提升整体的音频处理效率。
在本发明的另一实施例中，也可以是根据频率大小对补偿音频信号进行子带分解，得到多个子带信号；获取频率大于预设频率的子带信号作为目标子带信号；根据第二调制方法分别对各个目标子带信号的相位进行非线性调制，得到多个调制子带信号；将除目标子带信号外的子带信号和调制子带信号进行信号合成。其中，第二调制方法为使用第二调制公式对各个目标子带信号的相位进行非线性调制，第二调制公式包括第一调制参数，第一调制参数用于调整第一调制公式的非线性程度。
进一步的，在本发明的一种实施例中，第二调制公式为：
其中，α为第一调制参数，α的大小影响第一调制公式的非线性程度，H(z)为不同频率对应的需调整的相位大小，Z为频率，N为预设常数。
具体的，如图4所示为α设置为0.4时，第二调制公式的频率-相位响应图像的示意图。对大于预设频率的目标子带信号使用第二调制方法进行相位调制，不会对小于预设频率的部分子带信号产生影响，同时还可以减少相位调制过程的计算量，提升音频处理效率的同时，提升播放音频的保真度。
在本发明的另一实施例中，也可以是根据频率大小对补偿音频信号进行子带分解，得到多个子带信号；将频率大于预设频率的子带信号进行信号合成，得到第一合成信号；根据第二调制方法对第一合成信号的相位进 行非线性调制，得到第一调制合成信号；将除目标子带信号外的子带信号和第一调制合成信号进行信号合成。将将频率大于预设频率的子带信号进行信号合成，得到第一合成信号；根据第二调制方法对第一合成信号的相位进行非线性调制，可以减少相位调制过程中的计算量，提升整体的音频处理效率。
进一步的，前述步骤S201至步骤S204仅为本发明一种实施例中音频信号处理过程的举例说明，在本发明的另一种实施例中，音频信号处理过程还可以是如图5所示，包括以下步骤。
步骤S301：获取前音频输出信号。
步骤S302：根据前音频输出信号获取反馈信号。
步骤S303：根据反馈信号对音频输入信号进行信号补偿得到补偿音频信号。
可以理解的是，本实施例中的步骤S301至步骤S303与前一种实施例中的步骤S201至步骤S203大致相同，具体可以参照前述实施例的具体说明。
步骤S304：获取前补偿音频信号。
前补偿音频信号为前一采样时段的补偿音频信号。
步骤S305：根据前音频输出信号和前补偿音频信号计算得到反馈频响数据。
具体的，在发明的一种实施例中，可以获取前反馈频响数据，前反馈频响数据为前一采样时段的反馈频响数据；根据前音频输出信号、前补偿音频信号、以及前反馈频响数据计算得到当前采样时段的反馈频响数据。
进一步的，可以根据公式计算得到反馈频响数据。
其中，μ是迭代步长系数，W(n)＝[w0(n),w1(n),w2(n),...,wL-1(n)]T，w(n)为估计出来的反馈路径，n为采样时段,W(n)为前一采样时段的反馈频 响数据，W(n+1)为当前采样时段的反馈频响数据，R(n)＝[r(n),r(n-1),...,r(n-L+1)]T，r(n)为前一采样时段的补偿音频信号，r(n-L+1)为前L+2采样时段的补偿音频信号，L为预设的音频输出设备输出的音频反馈至音频输入设备的反馈路径长度。即为根据预设反馈路径长度L确定采样数量K，K＝L+2，获取与当前采样时段相邻的K个采样时段的音频输出信号形成音频输出阵列，R(n)即为音频输出阵列的逆转阵列，d(n)＝s(n)-z(n)，z(n)＝WT(n)R(n)，其中，s(n)为采样时段n的音频输入信号，z(n)为采样时段n的反馈信号，d(n)为采样时段n的补偿音频信号。使用上述公式即可以根据音频输出阵列、前补偿音频信号、以及前反馈频响数据计算得到当前采样时段的反馈频响数据。
步骤S306：判断反馈频响数据是否大于预设频响数据，若是，执行步骤S307，若否，执行步骤S308。
在本步骤中，预设频响数据为产生啸叫风险的门槛数据，反馈频响数据大于预设频响数据说明可能产生啸叫风险，此时执行步骤S306，反之反馈频响数据不大于预设频响数据说明不会产生啸叫风险，此时执行步骤S307。
步骤S307：对补偿音频信号的相位进行非线性调制。
步骤S308：将补偿音频信号作为音频输出信号。
在对补偿音频信号的相位进行非线性调制前先计算反馈频响数据，当反馈频响数据大于预设频响数据说明可能产生啸叫风险时才对补偿音频信号的相位进行非线性调制，反之则不对补偿音频信号的相位进行非线性调制，从而在保证不产生啸叫的前提下，减少对补偿音频信号的相位进行调制的次数，减少相位调制的计算量，提升音频信号处理的效率。
步骤S103：将音频输出信号经由音频输出设备进行播放。
与现有技术相比，本发明实施例一所提供的音频处理方法中，音频输入设备采集到音频输入信号后，根据前音频输出信号获取反馈信号，前音频输出信号即为前一采样时段经由音频输出设备进行播放的音频输出信号，根据前音频输出信号获取得到的反馈信号即为预估的前音频输出信号经由 音频输出设备进行播放后被音频输入设备重新采集的部分形成的信号，根据反馈信号对音频输入信号进行信号补偿，即可消除音频输入信号中所包含的音频输出设备播放的音频被重新采集的部分，如此得到的补偿音频信号即为实际的环境音频的音频信号，将补偿音频信号作为音频输出信号进行播放既可有效的消除可能产生的啸叫；此外，由于音频处理是一个连续的过程，当前采样时段得到的音频输出信号会在下一采样时刻作为前音频输出信号对下一采样时刻采集到的音频输入信号进行音频信号处理，对补偿音频信号的相位进行非线性调制，可以增大补偿音频信号与下一采样时刻采集到的音频输入信号的差异度，从而避免补偿音频信号与下一采样时刻采集到的音频输入信号的音频内容相似度过高、导致对下一采样时刻采集到的音频输入信号进行信号补偿时消除了过多的音频内容，从而减少信号补偿过程可能导致的音频失真，实现在消除啸叫的同时减少音频信号失真的效果。
本发明实施例二提供了一种音频处理装置，如图6所示，包括：音频输入模块601，音频输入模块601用于采集音频输入信号；处理模块602，处理模块602用于获取当前采样时段音频输入模块601采集的音频输入信号，对音频输入信号进行音频信号处理得到音频输出信号，音频信号处理至少包括：获取前音频输出信号，前音频输出信号为前一采样时段的音频输出信号，根据前音频输出信号获取反馈信号，根据反馈信号对音频输入信号进行信号补偿得到补偿音频信号，对补偿音频信号的相位进行非线性调制；音频输出模块603，音频输出模块603用于对音频输出信号进行播放；存储模块604，存储模块604用于存储前音频输出信号。
与现有技术相比，本发明实施例二所提供的音频处理装置中，音频输入模块601采集到音频输入信号后，处理模块602根据存储模块604中存储的前音频输出信号获取反馈信号，前音频输出信号为存储模块604中存储的前一采样时段经由音频输出模块603进行播放的音频输出信号，根据前音频输出信号获取得到的反馈信号即为预估的前音频输出信号经由音频输出模块603进行播放后被音频输入模块601重新采集的部分形成的信号，根据反馈信号对音频输入信号进行信号补偿，即可消除音频输入信号中所 包含的音频输出模块603播放的音频被重新采集的部分，如此得到的补偿音频信号即为实际的环境音频的音频信号，处理模块602将补偿音频信号作为音频输出信号进行播放既可有效的消除可能产生的啸叫；此外，由于音频处理是一个连续的过程，当前采样时段得到的音频输出信号会在下一采样时刻作为前音频输出信号对下一采样时刻采集到的音频输入信号进行音频信号处理，对补偿音频信号的相位进行非线性调制，可以增大补偿音频信号与下一采样时刻采集到的音频输入信号的差异度，从而避免补偿音频信号与下一采样时刻采集到的音频输入信号的音频内容相似度过高、导致对下一采样时刻采集到的音频输入信号进行信号补偿时消除了过多的音频内容，从而减少信号补偿过程可能导致的音频失真，实现在消除啸叫的同时减少音频信号失真的效果。
本发明实施例三涉及一种电子设备，如图7所示，包括：音频输入设备701、音频输出设备702、至少一个处理器703；以及，与至少一个处理器703通信连接的存储器704；其中，存储器704存储有可被至少一个处理器703执行的指令，指令被至少一个处理器703执行，以使至少一个处理器703能够执行上述各实施例中的音频处理方法。
其中，存储器和处理器采用总线方式连接，总线可以包括任意数量的互联的总线和桥，总线将一个或多个处理器和存储器的各种电路连接在一起。总线还可以将诸如外围设备、稳压器和功率管理电路等之类的各种其他电路连接在一起，这些都是本领域所公知的，因此，本文不再对其进行进一步描述。总线接口在总线和收发机之间提供接口。收发机可以是一个元件，也可以是多个元件，比如多个接收器和发送器，提供用于在传输介质上与各种其他装置通信的单元。经处理器处理的数据通过天线在无线介质上进行传输，进一步，天线还接收数据并将数据传送给处理器。
处理器负责管理总线和通常的处理，还可以提供各种功能，包括定时，外围接口，电压调节、电源管理以及其他控制功能。而存储器可以被用于存储处理器在执行操作时所使用的数据。
本发明实施例四涉及一种计算机可读存储介质，存储有计算机程序。计算机程序被处理器执行时实现上述方法实施例。
即，本领域技术人员可以理解，实现上述实施例方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，该程序存储在一个存储介质中，包括若干指令用以使得一个设备(可以是单片机，芯片等)或处理器(processor)执行本申请各个实施例方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-OnlyMemory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到的变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应以权利要求的保护范围为准。

Claims (13)

  1. 一种音频处理方法，其特征在于，应用于包括音频输入设备和音频输出设备的音频处理装置，所述方法包括：
    获取当前采样时段所述音频输入设备采集的音频输入信号，对所述音频输入信号进行音频信号处理得到音频输出信号，将所述音频输出信号经由所述音频输出设备进行播放；
    所述音频信号处理包括：获取前音频输出信号，所述前音频输出信号为前一采样时段的所述音频输出信号，根据所述前音频输出信号获取反馈信号，根据所述反馈信号对所述音频输入信号进行信号补偿得到补偿音频信号，对所述补偿音频信号的相位进行非线性调制。
  2. 根据权利要求1所述的方法，其特征在于，所述音频信号处理还包括：
    获取前补偿音频信号，所述前补偿音频信号为前一采样时段的补偿音频信号；
    根据所述前音频输出信号和所述前补偿音频信号计算得到反馈频响数据，判断所述反馈频响数据是否大于预设频响数据；
    若所述反馈频响数据大于预设频响数据，执行步骤：对所述补偿音频信号的相位进行非线性调制。
  3. 根据权利要求2所述的方法，其特征在于，所述根据所述前音频输出信号和所述前补偿音频信号计算得到反馈频响数据，包括：
    获取前反馈频响数据，所述前反馈频响数据为前一采样时段的所述反馈频响数据；
    根据所述前音频输出信号、所述前补偿音频信号、以及所述前反馈频响数据计算得到当前采样时段的所述反馈频响数据。
  4. 根据权利要求3所述的方法，其特征在于，所述根据所述前音频输出信号、所述前补偿音频信号、以及所述前反馈频响数据计算得到当前采样时段的所述反馈频响数据，包括：
    根据预设反馈路径长度确定采样数量K；
    获取与当前采样时段相邻的K个采样时段的所述音频输出信号形成音频输出阵列；
    根据所述音频输出阵列、所述补偿音频信号、以及所述前反馈频响数据计算得到当前采样时段的所述反馈频响数据。
  5. 根据权利要求1所述的方法，其特征在于，所述对所述补偿音频信号的相位进行非线性调制，包括：
    根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；
    根据第一调制方法分别对各个所述子带信号的相位进行非线性调制，得到多个调制子带信号；
    将所述多个调制子带信号进行信号合成。
  6. 根据权利要求1所述的方法，其特征在于，所述对所述补偿音频信号的相位进行非线性调制，包括：
    根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；
    分别对各个所述子带信号进行卷积和限幅处理；
    将所述卷积和限幅处理后的所述多个子带信号进行信号合成，得到合成信号；
    根据第一调制方法对所述合成信号的相位进行非线性调制。
  7. 根据权利要求5或6所述的方法，其特征在于，所述第一调制方法，包括：
    使用第一调制公式进行非线性调制；
    所述第一调制公式包括第一调制参数和第二调制参数，所述第一调制参数用于调整所述第一调制公式的非线性程度，所述第二调制参数用于调整所述第一调制公式的斜率。
  8. 根据权利要求1所述的方法，其特征在于，所述对所述补偿音频信号 的相位进行非线性调制，包括：
    根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；
    获取频率大于预设频率的所述子带信号作为目标子带信号；
    根据第二调制方法分别对各个所述目标子带信号的相位进行非线性调制，得到多个调制子带信号；
    将除所述目标子带信号外的所述子带信号和所述调制子带信号进行信号合成。
  9. 根据权利要求1所述的方法，其特征在于，所述对所述补偿音频信号的相位进行非线性调制，包括：
    根据频率大小对所述补偿音频信号进行子带分解，得到多个子带信号；
    获取频率大于预设频率的所述子带信号作为目标子带信号；
    将所述目标子带信号进行信号合成，得到第一合成信号；
    根据第二调制方法对第一合成信号的相位进行非线性调制，得到第一调制合成信号；
    将除所述目标子带信号外的所述子带信号和所述第一调制合成信号进行信号合成。
  10. 根据权利要求8或9所述的方法，其特征在于，所述第二调制方法，包括：
    使用第二调制公式进行非线性调制；
    所述第二调制公式包括第一调制参数，所述第一调制参数用于调整所述第二调制公式的非线性程度。
  11. 一种音频处理装置，其特征在于，包括：
    音频输入模块，所述音频输入模块用于采集音频输入信号；
    处理模块，所述处理模块用于获取当前采样时段所述音频输入模块采集的音频输入信号，对所述音频输入信号进行音频信号处理得到音频输出 信号，所述音频信号处理至少包括：获取前音频输出信号，所述前音频输出信号为前一采样时段的所述音频输出信号，根据所述前音频输出信号获取反馈信号，根据所述反馈信号对所述音频输入信号进行信号补偿得到补偿音频信号，对所述补偿音频信号的相位进行非线性调制；
    音频输出模块，所述音频输出模块用于对所述音频输出信号进行播放；
    存储模块，所述存储模块用于存储所述前音频输出信号。
  12. 一种电子设备，其特征在于，包括：
    音频输入设备、音频输出设备、至少一个处理器；以及，与所述至少一个处理器通信连接的存储器；
    其中，所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行如权利要求1至10中任意一项所述的音频处理方法。
  13. 一种计算机可读存储介质，存储有计算机程序，其特征在于，所述计算机程序被处理器执行实现权利要求1至10中任意一项所述的音频处理方法。
PCT/CN2023/103752 2022-11-04 2023-06-29 音频处理方法、装置、电子设备和计算机可读存储介质 WO2024093307A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211376944.XA CN115835093A (zh) 2022-11-04 2022-11-04 音频处理方法、装置、电子设备和计算机可读存储介质
CN202211376944.X 2022-11-04

Publications (1)

Publication Number Publication Date
WO2024093307A1 true WO2024093307A1 (zh) 2024-05-10

Family

ID=85526639

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/103752 WO2024093307A1 (zh) 2022-11-04 2023-06-29 音频处理方法、装置、电子设备和计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN115835093A (zh)
WO (1) WO2024093307A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115835093A (zh) * 2022-11-04 2023-03-21 恒玄科技(上海)股份有限公司 音频处理方法、装置、电子设备和计算机可读存储介质
CN117640868B (zh) * 2024-01-23 2024-08-16 宁波菊风系统软件有限公司 一种智能化双录系统及方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170223454A1 (en) * 2014-10-17 2017-08-03 Panasonic Intellectual Property Management Co., Ltd. Apparatus for removing howling and method for removing howling
JP2018125831A (ja) * 2017-02-03 2018-08-09 ミミー電子株式会社 遅延量可変によって信号周波数を変化させてハウリングを除去する補聴器回路
CN109862463A (zh) * 2018-12-26 2019-06-07 广东思派康电子科技有限公司 耳机语音回放方法、耳机及其计算机可读存储介质
CN111586512A (zh) * 2020-04-30 2020-08-25 歌尔科技有限公司 一种防啸叫方法、电子设备及计算机可读存储介质
CN111583949A (zh) * 2020-04-10 2020-08-25 南京拓灵智能科技有限公司 啸叫抑制的方法、装置和设备
CN115065921A (zh) * 2022-05-16 2022-09-16 武汉左点科技有限公司 一种防止助听器啸叫的方法及装置
CN115835093A (zh) * 2022-11-04 2023-03-21 恒玄科技(上海)股份有限公司 音频处理方法、装置、电子设备和计算机可读存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170223454A1 (en) * 2014-10-17 2017-08-03 Panasonic Intellectual Property Management Co., Ltd. Apparatus for removing howling and method for removing howling
JP2018125831A (ja) * 2017-02-03 2018-08-09 ミミー電子株式会社 遅延量可変によって信号周波数を変化させてハウリングを除去する補聴器回路
CN109862463A (zh) * 2018-12-26 2019-06-07 广东思派康电子科技有限公司 耳机语音回放方法、耳机及其计算机可读存储介质
CN111583949A (zh) * 2020-04-10 2020-08-25 南京拓灵智能科技有限公司 啸叫抑制的方法、装置和设备
CN111586512A (zh) * 2020-04-30 2020-08-25 歌尔科技有限公司 一种防啸叫方法、电子设备及计算机可读存储介质
CN115065921A (zh) * 2022-05-16 2022-09-16 武汉左点科技有限公司 一种防止助听器啸叫的方法及装置
CN115835093A (zh) * 2022-11-04 2023-03-21 恒玄科技(上海)股份有限公司 音频处理方法、装置、电子设备和计算机可读存储介质

Also Published As

Publication number Publication date
CN115835093A (zh) 2023-03-21

Similar Documents

Publication Publication Date Title
WO2024093307A1 (zh) 音频处理方法、装置、电子设备和计算机可读存储介质
JP6436934B2 (ja) 動的閾値を用いた周波数帯域圧縮
US9870783B2 (en) Audio signal processing
US9437182B2 (en) Active noise reduction method using perceptual masking
RU2626987C2 (ru) Устройство и способ для улучшения воспринимаемого качества воспроизведения звука путем объединения активного шумоподавления и компенсации воспринимаемого шума
JP5448771B2 (ja) 音響処理装置及び方法
CN100525101C (zh) 使用波束形成算法来记录信号的方法和设备
US10819305B2 (en) Method and device for adjusting sound quality
US8625813B2 (en) Asymmetric polynomial psychoacoustic bass enhancement
US8538052B2 (en) Generation of probe noise in a feedback cancellation system
CN115442709B (zh) 音频处理方法、虚拟低音增强系统、设备和存储介质
US11595764B2 (en) Tuning method, manufacturing method, computer-readable storage medium and tuning system
JP2014060691A (ja) 音声処理装置および方法、並びにプログラム
JP2004187165A (ja) 音声通信装置
JP2009175420A (ja) 音補正装置
CN111770404A (zh) 录音方法、装置、电子设备及可读存储介质
CN111629313B (zh) 包括环路增益限制器的听力装置
EP3595172B1 (en) Systems and methods for processing an audio signal for replay on an audio device
CN116405822A (zh) 一种应用于开放式蓝牙耳机的低音增强系统及方法
EP4021008B1 (en) Voice signal processing method and device
CN115714948A (zh) 一种音频信号处理方法、装置及存储介质
WO2021055513A1 (en) Mitigating acoustic feedback in hearing aids with frequency warping by all-pass networks
US11184715B1 (en) Hearing devices and methods for implementing an adaptively adjusted cut-off frequency
WO2024134805A1 (ja) 再生音補正装置、再生音補正方法、プログラム
US20230029841A1 (en) Perceptual bass extension with loudness management and artificial intelligence (ai)

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23884246

Country of ref document: EP

Kind code of ref document: A1