Audio processing methods and noise reduction devices

Abstract

This application discloses an audio processing method and a noise reduction device. The audio processing method and the noise reduction device obtain the sound signal of the sound emitted by the sound source at a first position, obtain a first sound signal, obtain the sound signal when the sound emitted by the sound source propagates to a second position, and obtain a second sound signal. The first position is the sound source position corresponding to the sound source, and the second position is within a preset distance range relative to the sound source position. Then, based on the first sound signal and the second sound signal, a noise reduction wave for noise reduction is generated, and the noise reduction wave is output at a third position (the third position and the second position satisfy a preset proximity condition) so as to reduce the sound signal of the sound source that has propagated to the third position as noise through the output noise reduction wave.

Description

Technical Field

[0001] This application relates to an audio processing method and a noise reduction device. Background Technology

[0002] Currently, most companies and enterprises have open-plan work environments, where different employees typically work in different locations within the open space (such as cubicles or different locations in non-cubicle office areas), such as working, studying, or holding meetings. Working in an open-plan environment can easily lead to mutual interference. The sounds produced by employees, such as talking or typing, can easily disturb those next to them, or they can be disturbed by the sounds produced by other employees next to them.

[0003] In addition, working from home is often done in an open-plan environment, where family members may intrude at any time or the noise generated by one's own work may disturb family members.

[0004] Existing noise reduction technologies are mainly applicable to situations where the location and sound frequency are relatively fixed, such as cars, air conditioners, and range hoods. They are basically ineffective for audio noise reduction in the above-mentioned or similar scenarios. Summary of the Invention

[0005] Therefore, this application discloses the following technical solution:

[0006] A noise reduction processing method, the method comprising:

[0007] Obtain the sound signal emitted by the sound source at the first position to obtain the first sound signal; the first position is the sound source position corresponding to the sound source;

[0008] Obtain the sound signal when the sound emitted by the sound source propagates to the second position, and obtain the second sound signal; the second position is within a preset distance range relative to the position of the sound source;

[0009] Based on the first sound signal and the second sound signal, a noise reduction wave is generated for noise reduction;

[0010] The noise-reduced wave is output at the third position so that the sound signal from the sound source that has traveled to the third position is used as noise for noise reduction; the third position and the second position satisfy a preset proximity condition.

[0011] Optionally, obtaining the sound signal emitted by the sound source at the first location and obtaining the first sound signal includes: obtaining the sound signal of the sound source collected by the first audio acquisition device located at the first location and obtaining the first sound signal;

[0012] The step of obtaining the sound signal when the sound emitted by the sound source propagates to the second position and obtaining the second sound signal includes: obtaining the sound signal of the sound source synchronously acquired by the second audio acquisition device set at the second position and obtaining the second sound signal.

[0013] Optionally, generating a noise-reducing wave for noise reduction based on the first sound signal and the second sound signal includes:

[0014] Generate an inverted sound wave whose corresponding signal waveform is the opposite phase of the signal waveform of the first sound signal;

[0015] The antiphase sound wave is adjusted according to the first sound signal and the second sound signal, and the adjusted antiphase sound wave is used as the noise reduction wave.

[0016] Optionally, generating an inverted sound wave whose corresponding signal waveform is the opposite phase to the signal waveform of the first sound signal includes:

[0017] The phase of the first sound signal waveform is reversed using hardware circuitry to obtain the inverted sound wave of the first sound signal.

[0018] Optionally, adjusting the antiphase sound wave according to the first sound signal and the second sound signal includes:

[0019] Determine the time difference between the second audio acquisition device and the first audio acquisition device synchronously acquiring the sound signal of the sound source;

[0020] Based on the time difference, determine the distance between the second position and the first position;

[0021] The waveform of the antiphase acoustic wave is phase-shifted based on the distance.

[0022] Optionally, generating a noise-reducing wave for noise reduction based on the first sound signal and the second sound signal includes:

[0023] Determine a first target sound signal in the first sound signal that matches a preset target voiceprint feature, and a second target sound signal in the second sound signal that matches the target voiceprint feature;

[0024] Based on the first target sound signal and the second target sound signal, a noise reduction wave is generated; the noise reduction wave is used to reduce the sound signal that the sound source has propagated to the third position and matches the target voiceprint features as noise.

[0025] Optionally, wherein the noise-reduced wave is output at the third position via an audio output device, the method further includes, before outputting the noise-reduced wave at the third position:

[0026] The third audio signal is obtained by acquiring the sound signal collected by the third audio acquisition device located at the output end of the audio output device;

[0027] Feedback information is determined based on the first sound signal and the third sound signal, and the noise reduction wave is adjusted based on the feedback information.

[0028] Optionally, determining feedback information based on the first sound signal and the third sound signal, and adjusting the noise reduction wave based on the feedback information, includes:

[0029] Based on the first sound signal and the third sound signal, determine the phase difference between the waveforms of the first sound signal and the third sound signal, and adjust the phase of the noise reduction wave based on the phase difference;

[0030] And / or, based on the first sound signal and the third sound signal, determine the amplitude difference between the sound signal from the sound source that has propagated to the third position and the noise-reduced wave, and adjust the amplitude of the noise-reduced wave based on the amplitude difference.

[0031] Optionally, the adjustment values ​​for the phase and / or amplitude of the noise-reduced wave are determined based on the phase difference and / or amplitude difference using a pre-built adjustment model.

[0032] A noise reduction device, comprising:

[0033] Memory, used to store at least one set of computer instructions;

[0034] A processor is configured to perform the following processing by invoking and executing the instruction set stored in the memory:

[0035] Obtain the sound signal emitted by the sound source at the first position to obtain the first sound signal; the first position is the sound source position corresponding to the sound source;

[0036] Obtain the sound signal when the sound emitted by the sound source propagates to the second position, and obtain the second sound signal; the second position is within a preset distance range relative to the position of the sound source;

[0037] Based on the first sound signal and the second sound signal, a noise reduction wave is generated for noise reduction;

[0038] The noise-reduced wave is output at the third position so that the sound signal from the sound source that has traveled to the third position is used as noise for noise reduction; the third position and the second position satisfy a preset proximity condition.

[0039] As can be seen from the above scheme, the audio processing method and noise reduction device disclosed in this application select a second and a third position that meet the proximity condition at the noise reduction end (such as the boundary between two adjacent workstations) where noise reduction is required. When noise reduction is needed for the sound emitted by the sound source at the noise reduction end, the first sound signal corresponding to the sound emitted by the sound source at the sound source position and the second sound signal corresponding to the sound emitted by the sound source when it propagates to the second position are obtained. A noise reduction wave is generated based on the first and second sound signals and output at the third position, thereby at least partially eliminating the sound signal of the sound source that has propagated to the third position. This achieves noise reduction at the noise reduction end by treating the sound signal of the sound source that has propagated to this position as noise. Therefore, for scenarios such as office work and self-study, the speaker's voice at a certain distance from the speaker's location (i.e., the noise reduction end) can be treated as noise and eliminated / reduced. Correspondingly, the sound of the speaker and other sound sources can be limited to the local area where the sound source is located (equivalent to providing an audio shield area for the sound source), which can prevent the speaker's voice from interfering with others in scenarios such as office work and self-study. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating one of the noise reduction processing methods provided in this application;

[0042] Figure 2 This is a schematic diagram of the noise reduction equipment deployment for fixed-point noise reduction in an office setting, as provided in this application.

[0043] Figure 3 This is another flowchart illustrating the noise reduction processing method provided in this application;

[0044] Figure 4 This is a circuit diagram of a noise reduction wave phase-modulated based on a modulation signal, provided in this application.

[0045] Figure 5 This is a schematic diagram of the noise reduction logic provided in this application;

[0046] Figures 6-7 These are schematic diagrams of the composition structure of the noise reduction device provided in this application. Detailed Implementation

[0047] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] This application discloses an audio processing method and a noise reduction device, applicable to, but not limited to, targeted noise reduction of different speakers' voices in scenarios such as office work and study, to prevent the voices of speakers or other audio sources from interfering with others or with each other, thus confining the voices of speakers or other audio sources to a local area where the audio source is located (equivalent to providing an audio shield area for the audio source). The audio processing method disclosed in this application can be applied to a noise reduction device, which has certain data processing / computing capabilities to support the operation of the audio processing method to achieve the required noise reduction processing. The noise reduction device will be described in detail in the following embodiments.

[0049] See Figure 1 The provided flowchart illustrates the noise reduction processing method, which includes the following steps:

[0050] Step 101: Obtain the sound signal emitted by the sound source at the first position to obtain the first sound signal.

[0051] The audio source can be, but is not limited to, a speaker in a setting such as an office or study room.

[0052] The first position refers to the location of the sound source, which can specifically refer to the actual location of the sound source, or a location that is close to or adjacent to the actual location of the sound source and can be approximated as the actual location of the sound source. Optionally, the first position being close to or adjacent to the actual location of the sound source can mean that the distance between the first position and the actual location of the sound source does not exceed a relatively small set distance value (e.g., 0.1m, 0.2m, etc.).

[0053] Specifically, one or more first audio acquisition devices can be set at the first position to acquire the sound signal emitted by the sound source at the first position as the first sound signal. This signal is essentially the original sound signal emitted by the sound source at the source end of the sound source.

[0054] For example, in an office setting, a microphone or a microphone array containing multiple microphones is set up within 0.1m of each workstation area, and the sound signal of the corresponding office worker is detected by the set microphone or microphone array, and the detected signal is used as the first sound signal.

[0055] In office settings, a common scenario is that office workers at different workstations communicate about work via phone calls or voice / video conferences as needed. During these calls or conferences, staff inevitably emit voice messages, which are usually only useful to themselves and useless to those at neighboring workstations. This can even disrupt or interfere with the normal work of those around them. The main purpose of this application is to address this scenario and similar situations by performing point-to-point noise reduction on the sounds emitted by office workers, thereby weakening or eliminating the sound signal at a specific location (e.g., the boundary between two adjacent workstations).

[0056] The first sound signal obtained in this step is mainly for the purpose of the above, and is used as one of the bases for generating noise reduction waves in the subsequent process.

[0057] Step 102: Obtain the sound signal when the sound emitted by the sound source propagates to the second position, and obtain the second sound signal; the second position is within a preset distance range relative to the position of the sound source.

[0058] Unlike the first position, which is the location of the sound source, the second position is essentially a location within the area corresponding to the noise reduction end that has noise reduction requirements. The area corresponding to the noise reduction end that has noise reduction requirements is the fixed-point location area determined in this application for fixed-point noise reduction. Taking the fixed-point noise reduction of the sound emitted by office workers in the above-mentioned office scenario as an example, preferably, the fixed-point location area can be the boundary area corresponding to the boundary of different workstations.

[0059] In practical applications, one or more second audio acquisition devices can be set at the second location. When the first audio acquisition device is used to acquire the first sound signal emitted by the sound source at the first location, the one or more second audio acquisition devices at the second location simultaneously acquire the sound signal emitted by the sound source when it propagates to that location (the second location) as the second sound signal. For example, a mic or a mic array containing multiple mics can be set at the boundary between different workstations. When the mic set within 0.1m of the office staff is used to acquire the first sound signal of the office staff at the sound source location, the mic set at the workstation boundary simultaneously acquires the second sound signal of the staff that has propagated to that boundary.

[0060] Preferably, the mic or mic array deployed at the second position is oriented toward the sound source that is the target of the acquisition, so as to better capture the sound source sound transmitted in the air.

[0061] The obtained second sound signal is also used as one of the bases for generating noise-reducing waves in the future.

[0062] Step 103: Generate a noise reduction wave for noise reduction based on the first sound signal and the second sound signal.

[0063] After obtaining the first and second sound signals, the first and second sound signals are further combined to generate a noise-reduced wave for noise reduction at the noise reduction end. This process can be implemented as follows:

[0064] 11) Generate an inverted sound wave that is the opposite phase of the signal waveform of the first sound signal.

[0065] Preferably, in this embodiment of the application, a hardware circuit is used to invert the phase of the signal waveform of the first sound signal. For example, a hardware amplifier circuit is directly used to invert the phase of the signal waveform of the first sound signal at the source end of the sound source to obtain the inverted sound wave of the first sound signal.

[0066] When performing point-to-point noise reduction in office or similar scenarios, the distance between the second position and the sound source position (first position) is usually close, resulting in a requirement for a short total noise reduction processing time. For example, in office or meeting scenarios, the distance between two people is usually only 2 to 3 meters, which means that the distance between the workstation boundary (representing the second position) and the office worker (representing the first position) is usually only 0.5 to 1.5 meters, resulting in a total noise reduction processing time of only 3 to 4 ms. This application reverses the phase of the first sound signal waveform through hardware circuitry, which can effectively reduce the processing time for obtaining the inverted waveform of the first sound signal, so as to better meet the requirement of a short noise reduction time.

[0067] 12) Adjust the antiphase sound wave according to the first sound signal and the second sound signal, and use the adjusted antiphase sound wave as a noise reduction wave for noise reduction.

[0068] Specifically, the time difference between the second audio acquisition device and the first audio acquisition device synchronously acquiring the sound signal of the sound source can be determined, and the distance between the second position and the first position can be determined based on the time difference and the speed of sound propagation in the air. Then, the waveform of the antiphase sound wave is phase-shifted based on the distance between the second position and the first position.

[0069] Specifically, the wavelength of the sound wave emitted by the sound source can be calculated based on the frequency of the sound wave emitted by the sound source. Based on the distance between the second position and the first position and the wavelength of the sound wave emitted by the sound source, the phase difference between the second sound signal and the first sound signal can be calculated. This phase difference includes the corresponding number of complete phase periods that differ between the second sound signal and the first sound signal, as well as the phase difference within one phase period. Based on this, the phase-shifted sound wave of the first sound signal is processed according to the calculated phase difference, so that the phase-shifted phase-shifted sound wave (i.e., the noise-reduced wave) can match the waveform characteristics of the sound signal that the sound source propagates to the noise reduction end as much as possible, thereby eliminating / attenuating the sound signal that the sound source propagates to the noise reduction end.

[0070] Step 104: Output the noise-reduced wave at the third position, so as to reduce the noise of the sound signal from the sound source that has been propagated to the third position as noise through the output noise-reduced wave.

[0071] The third position and the second position meet a preset proximity condition. Optionally, the proximity condition can specifically mean that the distance between the third position and the second position does not exceed a relatively small set distance value (e.g., 0.05m, 0.1m, etc.). In practical applications, this distance can be customized according to requirements.

[0072] Similar to the second position, the third position also belongs to the location area corresponding to the noise reduction end that has noise reduction requirements. In other words, this application selects two positions (the second position and the third position) that meet the proximity condition for the noise reduction end that has noise reduction requirements.

[0073] The purpose of selecting the second position is to set up a second audio acquisition device to acquire a second sound signal, which serves as one of the bases for generating the noise reduction wave. The purpose of selecting the third position is to set up an audio output device to output a noise reduction wave, thereby eliminating / attenuating the sound signal from the sound source that has traveled to this position (the third position). Constraining the second and third positions to satisfy the proximity condition—that is, there is a certain distance between the second and third positions (to ensure they are not the same position and do not overlap), and the distance between them does not exceed a set value—is intended to: firstly, ensure that the waveform of the second sound signal acquired at the second position is sufficiently similar to the waveform of the sound signal from the sound source that has traveled to the third position, so that the noise reduction wave generated based on the waveform characteristics of the second sound signal can effectively reduce / attenuate the sound signal from the sound source that has traveled to the third position; secondly, allow sufficient signal processing time based on the distance between the second and third positions to generate the noise reduction wave from the second sound signal obtained at the second position, and to output the noise reduction wave in a timely manner at the third position, thereby achieving noise reduction of the sound signal from the sound source that has traveled to the third position.

[0074] It is easy to understand that the first sound signal, the second sound signal, and the sound signal that the sound source has propagated to the third position are essentially synchronization signals when the same sound emitted by the sound source is propagated to different positions.

[0075] After generating a noise-reduced wave based on the first and second sound signals, a noise-reduced wave can be output using an audio output device (e.g., a speaker / horn) located at a third position. This noise-reduced wave is superimposed on the sound signal that has traveled from the sound source to this position, thereby eliminating / weakening the sound signal that has traveled from the sound source to this position, thus achieving the effect of point-to-point noise reduction of the sound source at the third position.

[0076] It should be noted that, compared to the third position, the second position is closer to the sound source, while the third position is relatively farther away from the sound source. See [link / reference needed]. Figure 2 This provides an example of a method for addressing the noise reduction needs of office worker A in an office setting. The method involves selecting a first, second, and third location, and deploying a first audio acquisition device, a second audio acquisition device, and an audio output device at each location. The first audio acquisition device comprises a microphone array consisting of individual microphones (mic1) shown in the diagram. The second audio acquisition device comprises a microphone array consisting of individual microphones (mic2) shown in the diagram. The audio output device comprises individual speaker (SPK) shown in the diagram. Each microphone (mic1), microphone (mic2), and speaker (SPK) corresponds to a specific microphone. By processing the sound signals acquired by the corresponding microphones (mic1 and mic2), a noise-reduced wave is generated. This noise-reduced wave is then output using the corresponding SPK. This effectively filters / attenuates the sound signal of office worker A, which travels through the air to the SPK location, preventing the sound of office worker A from spreading to adjacent office workers.

[0077] As can be seen from the above scheme, the audio processing method disclosed in this application selects a second and a third position that meet the proximity condition at the noise reduction end (such as the boundary between two adjacent workstations) where noise reduction is required. When noise reduction is needed for the sound emitted by the sound source at the noise reduction end, the method obtains the first sound signal corresponding to the sound source position and the second sound signal corresponding to the sound source when it propagates to the second position. A noise reduction wave is generated based on the first and second sound signals and output at the third position. This at least partially eliminates the sound signal of the sound source reaching the third position, thus achieving noise reduction at the noise reduction end by treating the sound signal of the sound source reaching this position as noise. Therefore, for scenarios such as office work and self-study, the speaker's voice at a certain distance from the speaker (i.e., the noise reduction end) can be treated as noise for elimination / reduction. Correspondingly, the sound of the speaker or other sound source can be limited to the local area of ​​the sound source (equivalent to providing an audio shield area for the sound source), which can prevent the speaker's voice from interfering with others in scenarios such as office work and self-study.

[0078] Furthermore, optionally, in one embodiment, when generating a noise reduction wave for noise reduction based on the first sound signal and the second sound signal, a noise reduction wave for targeted noise reduction of the target speaker's voice can be generated according to actual needs, combined with the voiceprint characteristics of the target speaker.

[0079] In this implementation, the process of generating a noise-reduced wave based on the first and second sound signals can be achieved as follows:

[0080] 21) Determine the first target sound signal in the first sound signal that matches the preset target voiceprint features, and the second target sound signal in the second sound signal that matches the target voiceprint features.

[0081] Among them, the target voiceprint feature is the voiceprint feature of the target speaker. In practical applications, the voiceprint feature of the target speaker can be set for targeted noise reduction according to the needs.

[0082] Based on the set target voiceprint features, the parts of the first and second voice signals that match the target voiceprint features are extracted, thus obtaining the signal parts of the first and second voice signals that belong to the target speaker.

[0083] 22) Generate a noise reduction wave based on the first target sound signal and the second target sound signal; the noise reduction wave is used to reduce the noise of the sound signal that the sound source has propagated to the third position and matches the target voiceprint features.

[0084] First, an inverted sound wave with the corresponding signal waveform opposite to that of the first target sound signal can be generated. Then, the inverted sound wave of the first target sound signal is adjusted according to the first target sound signal and the second target sound signal. The adjusted inverted sound wave is used as a noise reduction wave for targeted noise reduction of the target speaker. For a more detailed processing procedure, please refer to the above process of generating a noise reduction wave based on the first sound signal and the second sound signal. The only difference between this and the above process of generating a noise reduction wave based on the first sound signal and the second sound signal is that in this embodiment, the generation of the noise reduction wave is based on the first target sound signal and the second target sound signal.

[0085] Subsequently, the noise-reduced wave can be output using an audio output device such as a speaker located at the third position. The output noise-reduced wave can be used to selectively filter out the sound signal of the target speaker that has traveled to this location as noise.

[0086] This embodiment is more suitable for noise reduction scenarios with relatively stable sound sources that do not change frequently. For example, in an office setting, the staff at each workstation are relatively fixed and change infrequently. Therefore, for this scenario, the voiceprints of the staff at each workstation can be recorded in the noise reduction device deployed at each workstation to perform targeted noise reduction on the sound signals of the staff at that workstation. However, for scenarios such as meetings and self-study, the participants change relatively frequently, and the non-targeted noise reduction method not based on voiceprint features in this application can be preferred.

[0087] In one embodiment, optionally, see [link to previous document]. Figure 3 The flowchart of the noise reduction processing method shown in this application may further include the following processing:

[0088] Step 301: Obtain the sound signal collected by the third audio acquisition device set at the output end of the audio output device, and obtain the third sound signal.

[0089] This embodiment further includes a third audio acquisition device at the output end of the audio output device, such as... Figure 2 The mic3 is set at the output of each speaker and obtains the sound signal collected by the third audio acquisition device as the third sound signal.

[0090] This application aims to eliminate the sound signal that has traveled from the sound source to the location by outputting a noise-reduced wave through an audio output device at a third location. However, in practical applications, it is often difficult to completely eliminate the sound signal due to insufficient mining of potential influencing parameters or errors during signal processing. Based on this, this embodiment uses a third audio acquisition device located at the output end of the audio output device to acquire a third sound signal and provide feedback. By adjusting the noise-reduced wave based on the feedback, the noise reduction effect is further improved.

[0091] The third sound signal is essentially a residual signal generated due to incomplete noise elimination. It is the signal obtained by superimposing the noise-reduced wave with the sound signal from the sound source after propagation to the third position. The noise-reduced wave will be adjusted based on this signal in the future.

[0092] Step 302: Determine feedback information based on the first sound signal and the third sound signal, and adjust the noise reduction wave based on the feedback information.

[0093] The process of determining feedback information based on the first and third sound signals, and adjusting the noise reduction wave based on the feedback information, can be further implemented as any one or more of the following:

[0094] 31) Based on the first sound signal and the third sound signal, determine the phase difference between the waveforms of the first sound signal and the third sound signal, and adjust the phase of the noise reduction wave based on the phase difference.

[0095] Specifically, after determining the waveform phase difference between the first and third sound signals, this phase difference can be used as negative feedback input to a pre-constructed adjustment model. Based on the input phase difference (and possibly combined with the waveform phase difference between the first sound signal and the noise wave or the second sound signal), the adjustment model determines the phase adjustment value for the noise-reducing wave using its corresponding AI (Artificial Intelligence) strategy (such as a strategy based on waveform phase AI offset technology). This phase adjustment value includes both the adjustment amount and the adjustment direction (e.g., forward or backward shift). Then, the determined phase adjustment value can be superimposed on the noise wave to appropriately shift the carrier frequency of the noise wave forward or backward, so that the final output noise-reducing wave better adapts to the waveform characteristics of the sound signal that has propagated from the sound source to the third position, thereby achieving a higher degree of noise reduction for that sound signal.

[0096] Among them, such as Figure 4 As shown, a corresponding modulation signal can be generated based on the phase adjustment value to modulate the phase of the noise reduction wave.

[0097] 32) Based on the first sound signal and the third sound signal, determine the amplitude difference between the sound signal that the sound source reaches the third position after propagation and the noise reduction wave, and adjust the amplitude of the noise reduction wave according to the amplitude difference.

[0098] It is easy to understand that the sound signal from the sound source that reaches the third position after propagation is out of phase (or nearly out of phase) with the noise-reduced wave. Here, the difference in amplitude between the sound signal from the sound source that reaches the third position and the noise-reduced wave specifically refers to the difference in waveform amplitude in the amplitude direction without considering the difference in phase between the two, that is, only the difference in absolute value of waveform amplitude.

[0099] The amplitude difference between the sound signal from the sound source at the third position and the noise-reduced wave includes the numerical value and direction of the amplitude difference. The direction of the amplitude difference is used to characterize whether the waveform amplitude of the noise-reduced wave is insufficient or excessive compared to the waveform amplitude of the sound signal from the sound source at the third position. The numerical value of the amplitude difference characterizes the specific amplitude value of the insufficient or excessive amplitude.

[0100] Specifically, the waveform amplitude of the first sound signal can be directly used as the value of the difference between the waveform amplitude of the sound signal that the sound source reaches the third position and the waveform amplitude of the noise-reduced wave. At the same time, the direction (insufficient or excessive) of the waveform amplitude difference between the sound signal that the sound source reaches the third position and the noise-reduced wave can be compared with the waveform direction of the first sound signal and the waveform direction of the third sound signal.

[0101] Based on this, the determined amplitude difference can be used as negative feedback input to the adjustment model. The adjustment model, based on the input amplitude difference and its corresponding AI strategy, determines the amplitude adjustment value for the noise-reducing wave. This amplitude adjustment value also includes both the adjustment amount and the adjustment direction (e.g., increasing or decreasing the amplitude of the noise-reducing wave). Subsequently, the determined amplitude adjustment value can be further superimposed on the noise wave to achieve the effect of adjusting the amplitude of the noise wave's carrier frequency. The adjusted noise wave will then be superimposed again at the speaker output end with the waveform of the sound source propagating through the air, thereby further reducing the amplitude of the sound source at the later stages of the speaker output.

[0102] This application embodiment adjusts the noise sound wave based on the third sound signal collected at the noise sound wave output end, so that the adjusted noise sound wave can better adapt to the waveform characteristics of the sound signal that the sound source has propagated to this location (the third position), thereby further improving the point-to-point noise reduction effect of the sound source at this location.

[0103] The following provides an application example of the method of this application.

[0104] This example provides a specific implementation process for targeted noise reduction of office workers' voices in an office setting. Taking targeted noise reduction of the voice emitted by office worker A as an example, multiple noise reduction devices equipped with speakers and microphones are arranged around the speaker in an area 0.5m to 1.5m away (i.e., the area corresponding to the boundary between adjacent workstations). The device provides three sets of microphones, each consisting of multiple mic1, multiple mic2, and multiple mic3 microphones. mic1, mic2, and mic3 are used as the first, second, and third audio acquisition devices, respectively, while the speakers serve as the audio output device. The specific deployment framework is as follows: Figure 2As shown. It is easy to understand that the noise reduction device also includes a main body with data calculation / processing functions. The mic1 is connected to the main body of the device using a 0.5m to 1.5m cable, and the mic2 is placed on the main body of the device and faces the speaker to capture the speaker's voice transmitted through the air.

[0105] Based on the deployed audio acquisition and output devices, the point-to-point noise reduction process for the sound emitted by the speaker, office worker A, is as follows:

[0106] 41) Acquire the first sound signal of the speaker collected by mic1, and the second sound signal of the speaker collected synchronously by mic2;

[0107] 42) Estimate the straight-line distance between the speaker and the device (the main body of the noise reduction device) based on the capture time difference between the first sound signal and the second sound signal, and determine the phase difference between the first sound signal and the second sound signal based on the estimated distance;

[0108] 43) The waveform of the first sound signal is phase-inverted to obtain its antiphase sound wave, and the antiphase sound wave is phase-shifted based on the estimated phase difference. The phase-shifted antiphase sound wave is used as the noise reduction wave.

[0109] For details, please refer to [link / reference]. Figure 5 The noise reduction logic diagram shown illustrates that the noise reduction device inputs the first and second sound signals collected by mic1 and mic2 respectively into the calculated sound field unit. This unit directly uses a hardware amplification circuit to invert the phase of the waveform of the first sound signal and estimates the phase difference between the first and second sound signals. Then, the phase-inverted sound wave and the estimated phase difference are sent to the calculated ReserseElimination sound unit. This unit performs phase shifting on the inverted sound wave and outputs the phase-shifted waveform as the noise reduction wave through an amplifier and a speaker.

[0110] 44) Feedback is provided based on the first sound signal and the third sound signal collected by the speaker output mic3 to adjust the noise reduction wave.

[0111] Specifically, the Dynamic Calibration processing unit simultaneously obtains the first sound signal and the third sound signal collected by the speaker output mic3, compares the phase difference between the two, and / or determines the difference in waveform amplitude between the sound signal that the speaker has propagated to this point and the noise reduction wave based on the waveform of the two signals, and sends it as negative feedback to the calculate Reserse Elimination sound unit.

[0112] The Reserse Elimination sound unit uses the input phase difference as the basis for phase adjustment to determine the phase adjustment value for the noise-reduced wave, and superimposes it onto the noise-reduced wave waveform to achieve the effect of appropriately shifting the carrier frequency of the noise-reduced wave waveform forward or backward. And / or, the Reserse Elimination sound unit uses the input amplitude difference as the basis for amplitude adjustment to determine the amplitude adjustment value for the noise-reduced wave, and superimposes it onto the noise-reduced wave waveform to achieve the effect of appropriately adjusting the amplitude of the carrier frequency of the noise-reduced wave.

[0113] The adjusted noise reduction wave will be superimposed on the waveform of the sound source propagating through the air at the speaker output end, thereby achieving a better and more significant reduction in the amplitude of the sound source at the later stage of the speaker output end.

[0114] This application also discloses a noise reduction device that can be used to implement the above-described noise reduction processing method. The noise reduction device has the following structure: Figure 6 As shown, it includes at least:

[0115] Memory 10 is used to store at least one set of computer instructions;

[0116] Computer instruction sets can be implemented in the form of computer programs.

[0117] The processor 20 is configured to implement the method flow of the noise reduction processing method disclosed in any of the above embodiments by calling and executing the instruction set stored in the memory.

[0118] The processor 20 can be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices.

[0119] Optional, such as Figure 7 As shown, the noise reduction device may also include a first audio acquisition device 30, a second audio acquisition device 40, and an audio output device 50.

[0120] The first audio acquisition device 30 is positioned at a first location and acquires the sound signal emitted by the sound source at the first location to obtain a first sound signal. The second audio acquisition device 40 is positioned at a second location and acquires the sound signal emitted by the sound source as it propagates to the second location to obtain a second sound signal. The audio output device 50 is positioned at a third location and outputs a noise-reduced wave generated based on the first and second sound signals. The noise-reduced wave is used to treat the sound signal from the sound source that has propagated to the third location as noise for noise reduction.

[0121] The first position is the location of the sound source, the second position is within a preset distance range relative to the sound source location, and the third position satisfies the preset proximity condition with respect to the second position.

[0122] Additionally, optional, such as Figure 7 As shown, the noise reduction device may also include a third audio acquisition device 60, which is set at the output end of the audio output device 50 and acquires the sound signal at this location as a third sound signal, so that the noise reduction wave can be adjusted based on the third sound signal in the future, so as to achieve a better point-to-point noise reduction effect on the sound emitted by the sound source at the third position.

[0123] In summary, the audio processing method and noise reduction device disclosed in this application have at least the following technical advantages compared with the prior art:

[0124] a) A method was proposed to target noise reduction devices at specific locations based on human voice, so as to reduce the amplitude of the waveform transmitted by the speaker at a specific location, thereby achieving a targeted noise reduction effect on the sound emitted by the speaker.

[0125] b) By using second-order feedback methods based on phase adjustment and amplitude adjustment respectively, the noise reduction waveform can be optimized and adjusted in an ultra-short time.

[0126] c) By placing multiple microphones in different locations, an audio shield area surrounded by the device is formed, which significantly reduces the volume of sound transmitted from outside the area at a fixed point (the boundary of the audio shield). However, the volume of sound inside the area (i.e. inside the shield) is not affected and will not affect the normal speech communication of the speaker inside the area.

[0127] d) Waveform phase AI offset technology can be used to gradually adapt to the sound characteristics of the sound source in order to weaken the sound propagation intensity at a fixed point.

[0128] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the various embodiments can be referenced to each other.

[0129] For ease of description, the above systems or devices are described separately as various modules or units based on their functions. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware components.

[0130] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0131] Finally, it should be noted that in this document, relational terms such as first, second, third, and fourth are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0132] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A noise reduction processing method, the method comprising: Obtain the sound signal from the sound source acquired by the first audio acquisition device located at the first position, and obtain the first sound signal; The first position is the position of the sound source corresponding to the sound source; The second audio signal is obtained by synchronously acquiring the sound signal of the sound source by the second audio acquisition device located at the second position; the second position is within a preset distance range relative to the position of the sound source. Determine a first target sound signal in the first sound signal that matches a preset target voiceprint feature, and a second target sound signal in the second sound signal that matches the target voiceprint feature; Generate an inverted sound wave whose corresponding signal waveform is opposite to that of the first target sound signal; The anti-phase sound wave is adjusted according to the first target sound signal and the second target sound signal, and the adjusted anti-phase sound wave is used as a noise reduction wave; the noise reduction wave is used to reduce noise from the sound signal that the sound source has propagated to the third position and matches the target voiceprint features. The noise-reduced wave is output at the third position so that the sound signal from the sound source that has traveled to the third position is used as noise for noise reduction. The third position and the second position satisfy a preset proximity condition; The step of adjusting the anti-phase sound wave according to the first target sound signal and the second target sound signal includes: Determine the time difference between the second audio acquisition device and the first audio acquisition device synchronously acquiring the sound signal of the sound source; Based on the time difference, determine the distance between the second position and the first position; The waveform of the antiphase acoustic wave is phase-shifted based on the distance.

2. The method according to claim 1, wherein generating a corresponding signal waveform that is the opposite phase to the signal waveform of the first sound signal comprises: The phase of the first sound signal waveform is reversed using hardware circuitry to obtain the inverted sound wave of the first sound signal.

3. The method according to claim 1, wherein, The method further includes, prior to outputting the noise-reduced wave at the third position via an audio output device, the noise-reduced wave being output at the third position: The third audio signal is obtained by acquiring the sound signal collected by the third audio acquisition device located at the output end of the audio output device; Feedback information is determined based on the first sound signal and the third sound signal, and the noise reduction wave is adjusted based on the feedback information.

4. The method according to claim 3, wherein determining feedback information based on the first sound signal and the third sound signal, and adjusting the noise reduction wave based on the feedback information, comprises: Based on the first sound signal and the third sound signal, determine the phase difference between the waveforms of the first sound signal and the third sound signal, and adjust the phase of the noise reduction wave based on the phase difference; And / or, based on the first sound signal and the third sound signal, determine the amplitude difference between the sound signal from the sound source that has propagated to the third position and the noise-reduced wave, and adjust the amplitude of the noise-reduced wave based on the amplitude difference.

5. The method according to claim 4, wherein, The adjustment values ​​for the phase and / or amplitude of the noise-reduced wave are determined based on the phase difference and / or amplitude difference using a pre-built adjustment model.

6. A noise reduction device, comprising: Memory, used to store at least one set of computer instructions; A processor is configured to perform the following processing by invoking and executing the instruction set stored in the memory: The sound signal of the sound source is obtained by the first audio acquisition device set at the first position, and the first sound signal is obtained; the first position is the sound source position corresponding to the sound source; The second audio signal is obtained by synchronously acquiring the sound signal of the sound source by the second audio acquisition device located at the second position; the second position is within a preset distance range relative to the position of the sound source. A first target sound signal matching a preset target voiceprint feature in the first sound signal is identified, and a second target sound signal matching the target voiceprint feature in the second sound signal is identified; an anti-phase sound wave with a corresponding signal waveform that is opposite in phase to the signal waveform of the first target sound signal is generated; the anti-phase sound wave is adjusted according to the first target sound signal and the second target sound signal, and the adjusted anti-phase sound wave is used as a noise reduction wave; the noise reduction wave is used to reduce noise from the sound signal that the sound source has propagated to the third position and matches the target voiceprint feature. The noise-reduced wave is output at the third position so that the sound signal from the sound source that has traveled to the third position is used as noise for noise reduction. The third position and the second position satisfy a preset proximity condition; The step of adjusting the anti-phase sound wave according to the first target sound signal and the second target sound signal includes: Determine the time difference between the second audio acquisition device and the first audio acquisition device synchronously acquiring the sound signal of the sound source; Based on the time difference, determine the distance between the second position and the first position; The waveform of the antiphase acoustic wave is phase-shifted based on the distance.

Images