CN115914927A - Call noise reduction method and device and noise reduction earphone - Google Patents
Call noise reduction method and device and noise reduction earphone Download PDFInfo
- Publication number
- CN115914927A CN115914927A CN202211566260.6A CN202211566260A CN115914927A CN 115914927 A CN115914927 A CN 115914927A CN 202211566260 A CN202211566260 A CN 202211566260A CN 115914927 A CN115914927 A CN 115914927A
- Authority
- CN
- China
- Prior art keywords
- signal
- microphone
- feedback
- feedforward
- noise
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Images
Landscapes
- Soundproofing, Sound Blocking, And Sound Damping (AREA)
Abstract
The embodiment of the invention discloses a method and a device for reducing noise in a call and a noise reduction earphone. The conversation noise reduction method is applied to an earphone with a self-adaptive noise elimination function, the earphone comprises a feedforward microphone and a feedback microphone, and the conversation noise reduction method comprises the following steps: acquiring feedforward microphone signals acquired by a feedforward microphone and feedback microphone signals acquired by a feedback microphone; obtaining a wind noise identification result according to the feedforward microphone signal; and combining the feedforward microphone signal and the feedback microphone signal to obtain a voice signal according to a wind noise identification result. Through the mode, the embodiment of the invention can combine the feedforward microphone and the feedback microphone, when the wind noise is large, the signals collected by the feedforward microphone are eliminated, and when the wind noise is small, the signals collected by the feedforward microphone and the feedback microphone are mixed, and the self-adaptive noise elimination function is combined to realize noise reduction. And other microphones do not need to be additionally arranged, so that the hardware cost is reduced.
Description
Technical Field
The embodiment of the invention relates to the field of noise reduction of earphones, in particular to a method and a device for reducing noise of a call and a noise reduction earphone.
Background
In a noise scene, people often wear active noise reduction earphones to reduce the noise actually heard by human ears and achieve better hearing experience. A typical active noise reduction headphone includes an out-of-ear feedforward microphone and an in-ear feedback microphone. The feedforward microphone outside the ear is used for detecting the noise condition outside the ear, an electric signal is generated through feedforward noise reduction and is transmitted to the loudspeaker to generate an acoustic signal which is equal to the noise in the ear in amplitude and opposite to the noise in the ear in direction, and therefore the purpose of reducing the noise in the ear is achieved. Because the feedforward noise reduction effect is limited, the feedback microphone in the ear can be used for reducing noise through feedback, so that the residual noise in the ear is further reduced, and better noise reduction experience is achieved. In addition, the existing feedforward microphone and feedback microphone of the active noise reduction earphone can also be used for conversation, that is, in the occasion of voice conversation of a user, the noise influence in an uplink voice signal (namely, a voice signal sent to another conversation party) is suppressed through a processing algorithm.
The earphone inevitably meets the condition of wind noise in the use process, and the principle of wind noise generation is as follows: when wind meets an obstacle, turbulent flow (also called turbulent flow) is generated, the turbulent flow enables the air pressure near the cavity of the microphone to fluctuate, noise generated by the turbulent flow is amplified through resonance with an air column in the cavity of the microphone, and the amplified noise is picked up by the microphone to generate wind noise. Wind noise is not generated in the human ear, and is only generated at the microphone end, so that after the feedforward noise reduction is started, the wind noise can be strung in the human ear, and the experience is poor when the user listens to music. Meanwhile, wind noise also has influence on the call, so that the call definition is reduced. In order to reduce the influence of wind noise, the wind noise is firstly identified, and then the influence of the wind noise is reduced through some measures.
Disclosure of Invention
In order to solve the above technical problem, one technical solution adopted by the embodiment of the present invention is: the conversation noise reduction method is applied to a headset with an adaptive noise elimination function, the headset comprises a feedforward microphone and a feedback microphone, and the conversation noise reduction method comprises the following steps: acquiring a feedforward microphone signal acquired by the feedforward microphone and a feedback microphone signal acquired by the feedback microphone; obtaining a wind noise identification result according to the feedforward microphone signal and the feedback microphone signal; and combining the feedforward microphone signal and the feedback microphone signal to obtain a voice signal according to the wind noise identification result.
In some embodiments, said obtaining a wind noise identification result from said feedforward microphone signal and said feedback microphone signal comprises: obtaining a wind noise energy value according to the feedforward microphone signal and the feedback microphone signal; or obtaining a wind noise signal-to-noise ratio from the feedforward microphone signal.
In some embodiments, said obtaining a wind noise energy value from said feedforward microphone signal and said feedback microphone signal comprises: performing Fourier transform on the feedforward microphone signal and the feedback microphone signal respectively to obtain a feedforward microphone frequency domain signal and a feedback microphone frequency domain signal; performing inverse feedback filtering processing on the feedback microphone frequency domain signal to obtain an inverse feedback filtering signal; carrying out inverse feedforward filtering processing on the feedforward microphone frequency domain signal and the inverse feedback filtering signal to obtain an inverse mixed filtering signal; and calculating the ratio of the energy of the inverse feedback filtering signal to the energy of the inverse mixed filtering signal to obtain the wind noise energy value.
In some embodiments, the inverse feedback filtering process is implemented by the following formula:
FB invfb ＝FBmic*(1-H fb * G) Wherein, FB invfb For the inverse feedback filtered signal, FBmic is the feedback microphone frequency domain signal, H fb Starting the frequency response of a feedback filter used when the feedback noise reduction is carried out for the earphone at the current moment, wherein G is a transfer function from the loudspeaker to a feedback microphone; the inverse feedforward filtering process is realized by the following formula:
FB inv ＝FB invfb -FFmic*H ff * G, wherein, FB inv And FFmic is the frequency domain signal of the feedforward microphone, and Hff is the frequency response of a feedforward filter used when the feedforward noise reduction is started at the current moment of the earphone.
In some embodiments, said obtaining a wind noise signal to noise ratio from said feed forward microphone signal comprises: carrying out Fourier transform on the feedforward microphone signal to obtain a feedforward microphone frequency domain signal; and acquiring the wind noise signal-to-noise ratio according to the frequency domain signal of the feedforward microphone.
In some embodiments, the obtaining a speech signal according to the wind noise recognition result by combining the feedforward microphone signal and the feedback microphone signal includes: comparing the wind noise energy value with a first energy threshold value and a second energy threshold value respectively; if the wind noise energy value is smaller than the first energy threshold value, namely no wind noise exists, outputting the feedforward microphone signal and the feedback microphone signal; if the wind noise energy value is smaller than the second energy threshold value and larger than or equal to the first energy threshold value, outputting the feedforward microphone signal and the feedback microphone signal; if the wind noise energy value is larger than or equal to the second energy threshold value, outputting the feedback microphone signal; the second energy threshold is greater than the first energy threshold.
In some embodiments, the obtaining a speech signal by combining the feedforward microphone signal and the feedback microphone signal according to the wind noise recognition result further includes: comparing the wind noise signal-to-noise ratio with a first signal-to-noise ratio threshold; if the wind noise signal-to-noise ratio is greater than or equal to the first signal-to-noise ratio threshold, outputting the feedforward microphone signal and the feedback microphone signal; and if the wind noise signal-to-noise ratio is smaller than the first signal-to-noise ratio threshold value, outputting the feedback microphone signal.
In some embodiments, said outputting said feedback microphone signal comprises: canceling the feedforward microphone signal; causing the feedback microphone to output the feedback microphone signal.
In some embodiments, said cancelling said feed-forward microphone signal comprises: turning off the feed-forward microphone; or attenuating the feed forward microphone signal.
In order to solve the above technical problem, another technical solution adopted by the embodiment of the present invention is: provided is a call noise reduction device including: the signal acquisition module is used for acquiring feedforward microphone signals acquired by the feedforward microphone and feedback microphone signals acquired by the feedback microphone; the wind noise identification module is used for obtaining a wind noise identification result according to the feedforward microphone signal; and the voice acquisition module is used for combining the feedforward microphone signal and the feedback microphone signal to acquire a voice signal according to the wind noise identification result.
In order to solve the above technical problem, another technical solution adopted by the embodiment of the present invention is: there is provided a noise reducing headphone comprising: a feedforward microphone, a feedback microphone, a speaker, and a processor, wherein the feedforward microphone is communicatively coupled to the processor, the feedforward microphone being used for feedforward noise reduction; the feedback microphone is in communication connection with the processor and is used for feedback noise reduction; the processor is in communication connection with the speaker; the processor is configured to perform the call noise reduction method as described above.
The implementation mode of the invention has the beneficial effects that: different from the situation of the prior art, the embodiment of the invention can combine the feedforward microphone and the feedback microphone, eliminate the signals collected by the feedforward microphone when the wind noise is large, mix the signals collected by the feedforward microphone and the feedback microphone when the wind noise is small, and realize noise reduction by combining the self-adaptive noise elimination function. And other microphones do not need to be additionally arranged, so that the hardware cost is reduced.
Drawings
Fig. 1 is a schematic flow chart of a call noise reduction method according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a method for obtaining a wind noise identification result according to a feedforward microphone signal and a feedback microphone signal according to an embodiment of the present invention;
FIG. 3 is a schematic flow chart of another method for obtaining a wind noise identification result according to a feedforward microphone signal and a feedback microphone signal according to an embodiment of the present invention;
FIG. 4 is a schematic flow chart of obtaining a speech signal according to a wind noise recognition result by combining a feedforward microphone signal and a feedback microphone signal according to an embodiment of the invention;
FIG. 5 is a schematic flow chart of obtaining a speech signal according to a wind noise energy value by combining a feedforward microphone signal and a feedback microphone signal according to an embodiment of the invention;
FIG. 6 is a schematic flow chart of obtaining a speech signal according to the wind noise SNR by combining the feedforward microphone signal and the feedback microphone signal according to the embodiment of the invention;
FIG. 7 is a schematic flow chart of obtaining a second speech signal from a feedback microphone signal according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of a call noise reduction device according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of a noise reduction headphone according to an embodiment of the present invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications can be made by persons skilled in the art without departing from the spirit of the invention. All falling within the scope of the present invention.
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Active Noise Cancellation (ANC) is a method of identifying unwanted sound sources as Noise by generating an "anti-Noise" signal to cancel the original Noise, thereby canceling the Noise in real time. Modern noise reducing headsets, when passive noise reduction does not produce the desired result, typically employ this technique to cancel external noise sources, especially at lower frequencies (< 1 KHz). This is useful in some situations, such as when workers need to protect themselves from the continuous noise of a factory, or airline passengers need to filter out the noise of an engine.
The existing topology architecture of ANC headphones includes a feed-forward ANC architecture, a feedback ANC architecture and a hybrid ANC architecture, and the number and position of microphones in the headphones determine the ANC architecture of the headphones. Among these, with respect to feed forward ANC architectures, in such a topological architecture only one microphone is used per side earpiece, the microphone being placed outside the earpiece. The measurement data of the microphone is used as a noise reference signal of an ANC algorithm, so that the main noise signals can be well detected before reaching the ears of people, and the method is the biggest advantage of the design. The person wearing the headset, even without realizing its presence, eliminates this noise through an algorithm. However, due to the lack of a feedback loop in the system, the algorithm cannot guarantee that the noise cancellation work is real-time. This is a major drawback of feed forward ANC architectures, which configuration ensures cancellation of mid frequency noise signals (between 1-2 kHz). However, since the microphone is placed outside, the performance may be affected by wind noise
The feedback ANC architecture also uses only one microphone, but the microphone is placed closer to the user's ear within the headset. The main advantage of this topology is that the microphone can accurately hear the sound signal that is going to enter the user's ear, and there is a feedback loop to iteratively remove noise from the system. Although the low frequency performance of this configuration is good, the mid-frequency noise signal at frequencies between 1-2KHz cannot be attenuated as effectively as the feed-forward ANC architecture. Although this topology is not effective in removing the mid-high frequency noise signal, it is very effective in removing the predictable narrowband component of the primary noise signal. This is therefore one of the most common topologies when implementing ANC in a headset. The drawback of not being able to remove the mid-high frequency noise signal is that this topology does not cancel effectively for the high frequency components due to the phase shift introduced in the secondary path (the signal path is returned from the output of the ANC module all the way to its input). The microphones used in this topology need to have a flat group delay in frequency and the differences between the microphones should be as low as possible. This type of topology design is more suitable for bluetooth headsets that require wider ANC bandwidth and would like to sacrifice moderate noise cancellation performance for this purpose.
The hybrid ANC architecture combines a feedforward ANC architecture and a feedback ANC architecture, and such a topology is to combine the advantages of the two previous topologies to balance the feedforward ANC architecture and the feedback ANC architecture and the drawbacks thereof. So as to obtain wider ANC bandwidth on the premise of not sacrificing the noise elimination performance.
Based on the hybrid ANC architecture, the embodiment of the present invention provides a call noise reduction method, and the embodiment of the present invention is applied to an earphone with an adaptive noise cancellation function, where the earphone includes a feedforward microphone disposed at a position near the outside of the ear of an earphone housing for picking up ambient noise outside the ear, and a feedback microphone disposed at the front end of a speaker in the earphone for picking up in-ear noise. Fig. 1 shows a schematic flow diagram of the call noise reduction method, which includes the following steps:
step S100: acquiring a feedforward microphone signal acquired by a feedforward microphone and a feedback microphone signal acquired by a feedback microphone;
as mentioned above, the feedforward microphone is disposed near the ear of the earphone housing to pick up the ambient noise signal outside the ear, and the feedback microphone is disposed at the front end of the speaker to pick up the noise signal in the ear. Therefore, the embodiment of the application can acquire the feedforward microphone signal acquired by the feedforward microphone and the feedback microphone signal acquired by the feedback microphone as the basic signal for identifying the wind noise.
Step S200: obtaining a wind noise identification result according to the feedforward microphone signal and the feedback microphone signal;
in some embodiments, obtaining the wind noise identification result from the feedforward microphone signal and the feedback microphone signal includes the following steps, which are schematically shown in fig. 2:
step S211: respectively carrying out Fourier transform on the feedforward microphone signal and the feedback microphone signal to obtain a feedforward microphone frequency domain signal and a feedback microphone frequency domain signal;
specifically, after feedforward microphone signals collected by a feedforward microphone and feedback microphone signals collected by a feedback microphone are obtained, for convenience of subsequent calculation and processing of the signals, the feedforward microphone signals and the feedback microphone signals are converted into frequency domains through Fourier transform, and then feedforward microphone frequency domain signals FFmic and feedback microphone frequency domain signals FBmic are obtained respectively.
Step S212: carrying out inverse feedback filtering processing on the feedback microphone frequency domain signal to obtain an inverse feedback filtering signal;
specifically, the feedback microphone frequency domain obtained by the methodThe signal FBmic is subjected to inverse feedback filtering processing to obtain an inverse feedback filtering signal FB invfb Here, the inverse feedback filtering process is understood to be a process of restoring the frequency domain signal picked up by the feedback microphone to a state when the earphone is not turned on for feedback noise reduction.
In some embodiments, the inverse feedback filtering process is implemented by the following equation:
FB invfb ＝FBmic*(1-H fb *G)， (1)
wherein, FB invfb For the inverse feedback filtered signal, FBmic is the feedback microphone frequency domain signal, H fb And G is a transfer function from the loudspeaker to the feedback microphone.
Step S213: carrying out inverse feedforward filtering processing on the feedforward microphone frequency domain signal and the inverse feedback filtering signal to obtain an inverse mixed filtering signal;
specifically, the inverse feedback filtering result FB is obtained invfb Then, the inverse feedforward filtering signal needs to be further processed by inverse feedforward filtering in combination with the feedforward microphone frequency domain signal to obtain an inverse hybrid filtering processed signal FBinv. Since the inverse feedforward filtering process is further performed on the basis of the inverse feedback process, the inverse feedforward filtering process is understood to be a process of restoring the frequency domain signal picked up by the feedback microphone to a state when the hybrid noise reduction (including feedforward noise reduction and feedback noise reduction) is not turned on by the headphone. It should be noted that, this step does not perform inverse feedforward filtering processing on the feedforward microphone frequency-domain signal itself, because the feedforward microphone frequency-domain signal is generated outside the ear and is not affected by the active noise reduction, it is only necessary to consider the effect of the feedforward microphone frequency-domain signal in the ear on the feedback microphone frequency-domain signal.
The inverse feedforward filtering process is realized by the following formula:
FB inv ＝FB invfb -FFmic*H ff *G， (2)
wherein, FB inv For the inverse hybrid filtered signal, FFmic is the feedforward microphone frequency domain signal, hff is the earphone currentThe frequency response of a feedforward filter used in feedforward noise reduction is turned on at all times, and G is the transfer function from the loudspeaker to the feedback microphone.
It should be noted that the transfer function G from the speaker to the feedback microphone in the above equations (1) and (2) can be determined by collecting the speaker sound source signal and the feedback microphone signal picked up by the feedback microphone and calculating the corresponding relationship between the two signals. There may be two ways of calculating here: one is obtained by off-line calculation beforehand (i.e., determined by measurement in a laboratory), and the transfer function G obtained by off-line calculation beforehand can be called directly at the time of use, which is less time-consuming. Considering that different people have different wearing conditions of the earphones, the in-ear structures have some differences, and the coupling degrees of the earphones and the ears of different people are different, so that the acquired signals are also different, and therefore the signals can be determined in a statistical mode after the signal data of multiple people are acquired in advance, and the calculation accuracy is improved. The other calculation mode is real-time calculation, and the more accurate transfer function G can be calculated according to the coupling degree of the ears and the earphones of different people, so that the accuracy is relatively higher. Specifically, which way to calculate the transfer function G is adopted, a person skilled in the art can flexibly select the transfer function according to actual situations, and the method is not limited in detail here.
Step S214: and calculating the ratio of the energy of the inverse feedback filtering signal to the energy of the inverse mixed filtering signal to obtain the wind noise energy value.
In some embodiments, the ratio of the energy of the inverse feedback filtered signal and the energy of the inverse hybrid filtered signal is used as the wind noise energy value, i.e. the wind noise identification result.
The above embodiment provides a method for acquiring a wind noise identification result, and uses a wind noise energy value as a wind noise identification result. In addition, the embodiment of the present invention further provides another wind noise identification result obtaining method, and a flow chart thereof is shown in fig. 3, and includes the following steps:
step S221: fourier transform is carried out on the feedforward microphone signal to obtain a feedforward microphone frequency domain signal;
specifically, after the feedforward microphone signal collected by the feedforward microphone is obtained, in order to facilitate subsequent calculation and processing of the signal, the feedforward microphone signal is converted into a frequency domain through fourier transform, and then the feedforward microphone frequency domain signal FFmic is obtained.
Step S222: obtaining a wind noise signal-to-noise ratio according to the frequency domain signal of the feedforward microphone;
the Signal-to-Noise Ratio, known by the English name SNR or S/N (Signal-Noise Ratio), is also called Signal-to-Noise Ratio. Refers to the ratio of signal to noise in an electronic device or system. The signal refers to an electronic signal from the outside of the device to be processed by the device, the noise refers to an irregular extra signal (or information) which does not exist in the original signal generated after passing through the device, and the signal does not change along with the change of the original signal.
The simplest signal-to-noise ratio estimation method is as follows: after the feedforward microphone signal is subjected to Fourier transform to obtain a feedforward microphone frequency domain signal, a threshold value is set on the frequency spectrum of the feedforward microphone frequency domain signal, an effective signal Xs is arranged above the threshold value, a noise signal Xn is arranged below the threshold value, and the SNR can be estimated according to the following formula:
SNR = upper threshold sample value Xs 2 And/or the sub-threshold sample value Xn 2 And (3)
SNR is in dB, SNR (dB) =10 log10 (SNR).
In some embodiments, the wind noise signal-to-noise ratio is obtained from the feed-forward microphone frequency domain signal by the signal-to-noise ratio estimation method described above.
It should be noted that the wind noise signal-to-noise ratio may be used as a wind noise recognition result, but unlike the wind noise energy value, the noise signal used for calculating the wind noise energy value is only wind noise, and the noise signal used for calculating the wind noise signal-to-noise ratio includes wind noise and other noise.
Preferably, the embodiment of the invention combines two wind noise identification result acquisition methods. Of course, in other embodiments, either one of the two may be selected to obtain the wind noise identification result. Specifically, which way to obtain the wind noise recognition result is adopted, and a person skilled in the art can flexibly select the wind noise recognition result according to the actual situation, and is not specifically limited herein.
Step S300: and according to the wind noise identification result, combining the feedforward microphone signal and the feedback microphone signal to obtain a voice signal.
In some embodiments, obtaining the speech signal according to the wind noise recognition result by combining the feedforward microphone signal and the feedback microphone signal includes the following steps, as shown in fig. 4:
step S311: and comparing the wind noise energy value with a first energy threshold value and a second energy threshold value respectively.
Specifically, step S312, step S313 or step S314 is executed according to the comparison result between the wind noise energy value and the first energy threshold and the comparison result between the wind noise energy value and the second energy threshold.
Step S312: and if the wind noise energy value is smaller than the first energy threshold value, namely no wind noise exists, outputting a feedforward microphone signal and a feedback microphone signal.
Step S313: and if the wind noise energy value is smaller than the second energy threshold value and larger than or equal to the first energy threshold value, namely no wind noise exists, outputting a feedforward microphone signal and a feedback microphone signal.
Step S314: and if the wind noise energy value is larger than or equal to the second energy threshold value, outputting a feedback microphone signal.
After step S312 or step S313 is executed, the wind noise signal-to-noise ratio is compared with the first signal-to-noise ratio threshold, which is specifically as follows:
step S321: and comparing the wind noise signal-to-noise ratio with a first signal-to-noise ratio threshold value.
Step S322: if the wind noise signal-to-noise ratio is greater than or equal to the first signal-to-noise ratio threshold, outputting a feedforward microphone signal and a feedback microphone signal.
Specifically, a feedforward microphone signal and a feedback microphone signal are output, and the feedforward microphone signal and the feedback microphone signal are mixed to obtain a voice signal.
Step S323: and if the wind noise signal-to-noise ratio is smaller than the first signal-to-noise ratio threshold value, outputting a feedback microphone signal.
Specifically, the feedback microphone signal is output as a voice signal.
In some embodiments, if the wind noise energy value is greater than or equal to the second energy threshold value, or the wind noise signal-to-noise ratio is less than the first signal-to-noise ratio threshold value, the method outputs the feedback microphone signal, specifically including the following steps, as shown in fig. 7:
step S331: the feed-forward microphone signal is cancelled.
Methods of canceling the feedforward microphone signal include, but are not limited to, turning off the feedforward microphone and attenuating the feedforward microphone signal. Preferably, embodiments of the present invention eliminate the feedforward microphone signal by turning off the feedforward microphone.
Step S332: causing the feedback microphone to output a feedback microphone signal.
In other embodiments, the method for obtaining the speech signal by recognizing the wind noise energy value and combining the feedforward microphone signal and the feedback microphone signal only uses the wind noise energy value as the wind noise recognition result, includes the following steps, as shown in fig. 5:
step S311: and comparing the wind noise energy value with a first energy threshold value and a second energy threshold value respectively.
Specifically, step S312, step S313 or step S314 is executed according to the comparison result between the wind noise energy value and the first energy threshold and the comparison result between the wind noise energy value and the second energy threshold.
Step S312: and if the wind noise energy value is smaller than the first energy threshold value, namely no wind noise exists, outputting a feedforward microphone signal and a feedback microphone signal.
Specifically, a feedforward microphone signal and a feedback microphone signal are output, and the feedforward microphone signal and the feedback microphone signal are mixed to obtain a voice signal.
Step S313: and if the wind noise energy value is smaller than the second energy threshold value and larger than or equal to the first energy threshold value, outputting a feedforward microphone signal and a feedback microphone signal.
Specifically, a feedforward microphone signal and a feedback microphone signal are output, and the feedforward microphone signal and the feedback microphone signal are mixed to obtain a voice signal.
Step S314: and if the wind noise energy value is larger than or equal to the second energy threshold value, outputting a feedback microphone signal.
Specifically, outputting the feedback microphone signal as a voice signal includes the following steps, as shown in fig. 7:
step S331: the feed-forward microphone signal is cancelled.
Methods of canceling the feedforward microphone signal include, but are not limited to, turning off the feedforward microphone and attenuating the feedforward microphone signal. Preferably, embodiments of the present invention eliminate the feedforward microphone signal by turning off the feedforward microphone.
Step S332: causing the feedback microphone to output a feedback microphone signal.
In other embodiments, the method for obtaining the speech signal by recognizing the wind noise signal-to-noise ratio and combining the feedforward microphone signal and the feedback microphone signal only uses the wind noise signal-to-noise ratio as the wind noise recognition result, includes the following steps, as shown in fig. 6:
step S321: and comparing the wind noise signal-to-noise ratio with a first signal-to-noise ratio threshold value.
Step S322: and if the wind noise signal-to-noise ratio is greater than or equal to the first signal-to-noise ratio threshold value, outputting a feedforward microphone signal and a feedback microphone signal.
Specifically, a feedforward microphone signal and a feedback microphone signal are output, and the feedforward microphone signal and the feedback microphone signal are mixed to obtain a voice signal.
Step S323: and if the wind noise signal-to-noise ratio is smaller than the first signal-to-noise ratio threshold value, outputting a feedback microphone signal.
Specifically, outputting the feedback microphone signal as a voice signal includes the following steps, as shown in fig. 7:
step S331: the feed-forward microphone signal is cancelled.
Methods of canceling the feedforward microphone signal include, but are not limited to, turning off the feedforward microphone and attenuating the feedforward microphone signal. Preferably, embodiments of the present invention eliminate the feedforward microphone signal by turning off the feedforward microphone.
Step S332: causing the feedback microphone to output a feedback microphone signal.
Different from the prior art, the embodiment of the invention can combine the feedforward microphone and the feedback microphone, eliminate the signals collected by the feedforward microphone when the wind noise is large, mix the signals collected by the feedforward microphone and the feedback microphone when the wind noise is small, and realize noise reduction by combining the self-adaptive noise elimination function. And other microphones do not need to be additionally arranged, so that the hardware cost is reduced.
Based on the above call noise reduction method, an embodiment of the present invention further provides a call noise reduction device, a schematic structural diagram of which is shown in fig. 8, and the device includes a signal acquisition module 100, a wind noise recognition module 200, and a voice acquisition module 300.
The signal acquiring module 100 is configured to acquire a feedforward microphone signal acquired by a feedforward microphone and a feedback microphone signal acquired by a feedback microphone;
the wind noise identification module 200 is configured to obtain a wind noise identification result according to the feedforward microphone signal;
the voice obtaining module 300 is configured to obtain a voice signal according to the wind noise recognition result by combining the feedforward microphone signal and the feedback microphone signal.
Based on the above call noise reduction method, an embodiment of the present invention further provides a noise reduction earphone, a schematic structural diagram of which is shown in fig. 9, where the noise reduction earphone includes a feedforward microphone 400, a feedback microphone 500, a speaker 600, and a processor 700. Wherein the feedforward microphone 40 is communicatively coupled to the processor 600; the feedback microphone 500 is communicatively connected to the processor 600; the processor 600 is communicatively coupled to the speaker 700, and the processor 600 is configured to perform the above-mentioned call noise reduction method.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes performed by the present specification and drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (11)
1. A conversation noise reduction method is applied to earphones with an adaptive noise elimination function, the earphones comprise a feedforward microphone and a feedback microphone, and the conversation noise reduction method comprises the following steps:
acquiring a feedforward microphone signal acquired by the feedforward microphone and a feedback microphone signal acquired by the feedback microphone;
obtaining a wind noise identification result according to the feedforward microphone signal and the feedback microphone signal;
and according to the wind noise identification result, combining the feedforward microphone signal and the feedback microphone signal to obtain a voice signal.
2. The method of claim 1, wherein obtaining a wind noise identification result from the feedforward microphone signal and the feedback microphone signal comprises:
obtaining a wind noise energy value according to the feedforward microphone signal and the feedback microphone signal; or
And acquiring a wind noise signal-to-noise ratio according to the feedforward microphone signal.
3. The method of claim 2, wherein obtaining a wind noise energy value from the feedforward microphone signal and the feedback microphone signal comprises:
performing Fourier transform on the feedforward microphone signal and the feedback microphone signal respectively to obtain a feedforward microphone frequency domain signal and a feedback microphone frequency domain signal;
performing inverse feedback filtering processing on the feedback microphone frequency domain signal to obtain an inverse feedback filtering signal;
carrying out inverse feedforward filtering processing on the feedforward microphone frequency domain signal and the inverse feedback filtering signal to obtain an inverse mixed filtering signal;
and calculating the ratio of the energy of the inverse feedback filtering signal to the energy of the inverse mixed filtering signal to obtain the wind noise energy value.
4. The method of claim 3, the headset further comprising a speaker, wherein the inverse feedback filtering process is implemented by the following equation:
FB invfb ＝FBmic*(1-H fb *G)，
wherein, FB invfb For the inverse feedback filtered signal, FBmic is the feedback microphone frequency domain signal, H fb Starting the frequency response of a feedback filter used when the feedback noise reduction is carried out for the earphone at the current moment, wherein G is the transfer function from the loudspeaker to the feedback microphone;
the inverse feedforward filtering process is realized by the following formula:
FB inv ＝FB invfb -FFmic*H ff *G，
wherein, FB inv And for the inverse mixed filtering signal, FFmic is the frequency domain signal of the feedforward microphone, and Hff is the frequency response of a feedforward filter used when the feedforward noise reduction is started at the current moment of the earphone.
5. The method of claim 2, wherein obtaining a wind noise signal-to-noise ratio from the feed-forward microphone signal comprises:
carrying out Fourier transform on the feedforward microphone signal to obtain a feedforward microphone frequency domain signal;
and acquiring the wind noise signal-to-noise ratio according to the frequency domain signal of the feedforward microphone.
6. The method of claim 3, wherein the combining the feedforward microphone signal and the feedback microphone signal to obtain a speech signal according to the wind noise recognition result comprises:
comparing the wind noise energy value with a first energy threshold value and a second energy threshold value respectively;
if the wind noise energy value is smaller than the first energy threshold value, namely no wind noise exists, outputting the feedforward microphone signal and the feedback microphone signal;
if the wind noise energy value is smaller than the second energy threshold value and larger than or equal to the first energy threshold value, outputting the feedforward microphone signal and the feedback microphone signal;
if the wind noise energy value is larger than or equal to the second energy threshold value, outputting the feedback microphone signal;
the second energy threshold is greater than the first energy threshold.
7. The method of claim 5, wherein the combining the feedforward microphone signal and the feedback microphone signal to obtain a speech signal according to the wind noise recognition result, further comprising:
comparing the wind noise signal-to-noise ratio with a first signal-to-noise ratio threshold;
if the wind noise signal-to-noise ratio is greater than or equal to the first signal-to-noise ratio threshold, outputting the feedforward microphone signal and the feedback microphone signal;
and if the wind noise signal-to-noise ratio is smaller than the first signal-to-noise ratio threshold value, outputting the feedback microphone signal.
8. The method of claim 6 or 7, wherein the outputting the feedback microphone signal comprises:
canceling the feedforward microphone signal;
causing the feedback microphone to output the feedback microphone signal.
9. The method of claim 8, wherein canceling the feedforward microphone signal comprises:
turning off the feed-forward microphone; or
Attenuating the feedforward microphone signal.
10. A speech noise reduction apparatus, comprising:
the signal acquisition module is used for acquiring feedforward microphone signals acquired by the feedforward microphone and feedback microphone signals acquired by the feedback microphone;
the wind noise identification module is used for obtaining a wind noise identification result according to the feedforward microphone signal;
and the voice acquisition module is used for combining the feedforward microphone signal and the feedback microphone signal to acquire a voice signal according to the wind noise identification result.
11. A noise reducing headset, comprising: a feedforward microphone, a feedback microphone, a speaker, and a processor, wherein,
the feedforward microphone is in communication connection with the processor and is used for feedforward noise reduction;
the feedback microphone is in communication connection with the processor and is used for feedback noise reduction;
the processor is in communication connection with the speaker;
the processor is configured to perform the method of call noise reduction according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211566260.6A CN115914927A (en) | 2022-12-07 | 2022-12-07 | Call noise reduction method and device and noise reduction earphone |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211566260.6A CN115914927A (en) | 2022-12-07 | 2022-12-07 | Call noise reduction method and device and noise reduction earphone |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115914927A true CN115914927A (en) | 2023-04-04 |
Family
ID=86486647
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211566260.6A Pending CN115914927A (en) | 2022-12-07 | 2022-12-07 | Call noise reduction method and device and noise reduction earphone |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115914927A (en) |
-
2022
- 2022-12-07 CN CN202211566260.6A patent/CN115914927A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107533838B (en) | Voice sensing using multiple microphones | |
| JP6017825B2 (en) | A microphone and earphone combination audio headset with means for denoising proximity audio signals, especially for "hands-free" telephone systems | |
| US11812217B2 (en) | Wearable audio device with inner microphone adaptive noise reduction | |
| Liebich et al. | Signal processing challenges for active noise cancellation headphones | |
| CN109195042B (en) | Low-power-consumption efficient noise reduction earphone and noise reduction system | |
| WO2022052244A1 (en) | Earphone speech activity detection method, earphones, and storage medium | |
| CN108712703A (en) | The high-efficient noise-reducing earphone and noise reduction system of low-power consumption | |
| US11743662B2 (en) | Voice input/output apparatus, hearing aid, voice input/output method, and voice input/output program | |
| CN112954530B (en) | Earphone noise reduction method, device and system and wireless earphone | |
| CN108847250B (en) | Directional noise reduction method and system and earphone | |
| CN111935584A (en) | Wind noise processing method and device for wireless earphone assembly and earphone | |
| Shen et al. | A wireless reference active noise control headphone using coherence based selection technique | |
| EP4118648A1 (en) | Audio processing using distributed machine learning model | |
| WO2021016001A1 (en) | Input signal-based frequency domain adaptive filter stability control | |
| JP2023040244A (en) | Earphone, voice processing method, and voice processing program | |
| US11889268B2 (en) | Method for operating a hearing aid system having a hearing instrument, hearing aid system and hearing instrument | |
| US11533555B1 (en) | Wearable audio device with enhanced voice pick-up | |
| CN115914927A (en) | Call noise reduction method and device and noise reduction earphone | |
| KR20050119758A (en) | Hearing aid having noise and feedback signal reduction function and signal processing method thereof | |
| JP5383008B2 (en) | Speech intelligibility improvement system and speech intelligibility improvement method | |
| US11074903B1 (en) | Audio device with adaptive equalization | |
| Spriet et al. | Combined feedback and noise suppression in hearing aids | |
| US11948546B2 (en) | Feed-forward adaptive noise-canceling with dynamic filter selection based on classifying acoustic environment | |
| WO2022198538A1 (en) | Active noise reduction audio device, and method for active noise reduction | |
| Akhtar et al. | Lattice adaptive filtering-based method for acoustic feedback cancellation in hearing aids with robustness against sudden changes in the feedback path |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |