Active noise reduction method, device, chip, earphone and storage medium

Abstract

The application discloses an active noise reduction method and device, a chip, earphones and a storage medium, and relates to the technical field of audio processing. The method comprises the following steps: a first processor determines a target active noise reduction mode according to the noise characteristics of an environment in which the earphones are located, and generates filter parameters; and a second processor performs active noise reduction processing based on the filter parameters provided by the first processor. According to the embodiment scheme, the ANC mode is intelligently determined based on the noise characteristics of environmental audio through the double processors, and the quality of active noise reduction of environmental noise in different scenes is improved.

Description

Technical Field

[0001] This application relates to the field of audio processing technology, and in particular to an active noise cancellation method, device, chip, earphone, and storage medium. Background Technology

[0002] With the development of audio technology, active noise cancellation has become the mainstream noise cancellation method for headphone devices compared to passive noise cancellation based on physical means. Active noise-canceling headphones collect ambient noise through microphones and use filters in the ANC (Active Noise Control) structure to determine the noise signal. Then, they play an anti-noise signal with the same amplitude and opposite phase as the noise signal through speakers, thus achieving noise cancellation.

[0003] In related technologies, active noise-canceling headphones determine the ANC mode based on the terminal device control or their own default settings. In a fixed ANC mode, the headphone device can only achieve one noise reduction effect through one filtering method, such as eliminating all ambient audio or retaining only human voices. When the user's environment changes, the headphone device cannot flexibly adapt to the characteristics of ambient noise and adjust the filtering method, making it difficult to meet the noise reduction effect needs of users when switching usage scenarios. Summary of the Invention

[0004] This application provides an active noise cancellation method, device, chip, earphone, and storage medium. It can intelligently determine the ANC mode based on noise characteristics through dual-processor (Digital Signal Processing, DSP), improving the quality of active noise cancellation for environmental noise in different scenarios. The technical solution is as follows:

[0005] On one hand, embodiments of this application provide an active noise cancellation method for headphones, the headphones including a first processor and a second processor, the method including:

[0006] The first processor determines the target active noise cancellation mode and generates filter parameters based on the noise characteristics of the environment in which the headphones are located;

[0007] The second processor performs active noise reduction processing based on the filter parameters provided by the first processor.

[0008] On the other hand, embodiments of this application provide an active noise cancellation device, the device comprising:

[0009] The first processor is used to determine the target active noise cancellation mode and generate filter parameters based on the noise characteristics of the environment in which the headphones are located.

[0010] The second processor is used to perform active noise reduction processing based on the filter parameters provided by the first processor.

[0011] On the other hand, embodiments of this application provide a chip, the chip including a first processor and a second processor, wherein the first processor is configured to:

[0012] The target active noise cancellation mode is determined based on the noise characteristics of the environment in which the headphones are located, and filter parameters are generated.

[0013] The second processor is configured as follows:

[0014] Active noise reduction processing is performed based on the filter parameters provided by the first processor.

[0015] On the other hand, embodiments of this application provide an earphone, the earphone including a processor and a memory, the processor including at least a first processor and a second processor, the memory storing at least a program, the at least a program being loaded and executed by the processor to implement the active noise cancellation method as described above.

[0016] On the other hand, embodiments of this application provide a computer-readable storage medium storing at least one program, wherein the at least one instruction is loaded and executed by a processor to implement the active noise reduction method as described above.

[0017] On the other hand, embodiments of this application provide a computer program product including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the active noise reduction method described above.

[0018] In this embodiment, the headphone device implements noise reduction through dual processors. The first processor determines a more suitable target active noise reduction mode from multiple active noise reduction modes based on the noise characteristics of the environment and generates corresponding filter parameters. The second processor performs active noise reduction processing based on these filter parameters. This solution analyzes the noise characteristics of the ambient audio and intelligently determines the target noise reduction mode that meets the user's needs from multiple active noise reduction modes suitable for different noise environments. Compared with related technologies, which only use one active noise reduction mode for noise reduction processing, this improves the quality of active noise reduction for environmental noise in different scenarios. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This invention provides a structural block diagram of an earphone according to an exemplary embodiment of the present application.

[0021] Figure 2 A flowchart of an active noise reduction method provided in an exemplary embodiment of this application is shown;

[0022] Figure 3 This illustration shows a schematic diagram of active noise reduction for target determination provided in an exemplary embodiment of this application;

[0023] Figure 4 A flowchart of an active noise reduction method provided in another exemplary embodiment of this application is shown;

[0024] Figure 5 This invention illustrates a process diagram of noise processing for a first processor provided in an exemplary embodiment of this application;

[0025] Figure 6 A structural diagram of a feedforward standard noise reduction mode provided in an exemplary embodiment of this application is shown;

[0026] Figure 7 A structural diagram of a hybrid standard noise reduction mode provided in an exemplary embodiment of this application is shown;

[0027] Figure 8 A structural diagram of an adaptive noise reduction mode provided in an exemplary embodiment of this application is shown;

[0028] Figure 9 A structural diagram of a directional noise reduction mode provided in an exemplary embodiment of this application is shown;

[0029] Figure 10 This invention illustrates a process diagram of noise processing by a second processor provided in an exemplary embodiment of this application.

[0030] Figure 11 A flowchart of a directional noise reduction method provided in an exemplary embodiment of this application is shown;

[0031] Figure 12 A system block diagram of beamforming provided in an exemplary embodiment of this application is shown;

[0032] Figure 13 A schematic diagram of a beamforming system provided in an exemplary embodiment of this application is shown;

[0033] Figure 14 A structural block diagram of an active noise cancellation device provided in an exemplary embodiment of this application is shown. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0035] For ease of understanding, the terms used in the embodiments of this application will be explained below.

[0036] Active Noise Control (ANC): Also known as active noise cancellation, it's a method to reduce the impact of external noise on headphone performance. Devices with ANC include a microphone, a processing chip, and a speaker. The device uses a microphone to pick up noise, and the processing chip determines the waveform characteristics of the noise. Based on this waveform, the chip calculates an anti-noise signal that is out of phase with the noise waveform, which is then played through the speaker. When ambient noise or built-in noise and the anti-noise signal are simultaneously transmitted to the ear, the noise signal and the inverse anti-noise signal cancel each other out, thus achieving noise reduction. Active noise cancellation is very effective at handling low-frequency noise, and compared to passive noise cancellation, users don't need to increase the audio volume to isolate external noise, reducing potential harm to the ears while achieving noise reduction.

[0037] Beamforming, also known as beamforming or spatial filtering, is a signal processing technique that transmits and receives radio or sound waves in a directional manner. It is commonly used in radar, sonar systems, wireless communications, acoustics, and biomedical devices. In audio processing, a beamformer, based on a defined sound source, adjusts the basic unit parameters of a phase array to perform constructive interference on signals at certain angles (target directions) and destructive interference on signals at other angles (non-target directions). This is followed by weighted summation and filtering of the signals output from each microphone in the microphone array—essentially merging multiple microphone signals—to ultimately output an audio signal in the desired direction.

[0038] Please refer to Figure 1 The diagram illustrates a structural block diagram of a headset provided in an exemplary embodiment of this application. The headset 100 may include one or more components such as an audio subsystem (Audio SS) 110 and a codec 120.

[0039] The audio subsystem 110 may include a digital signal processor (DSP) 111, a serial peripheral interface (SPI) 112, an integrated circuit built-in audio bus 113 (I2S, Inter-IC Sound), and a gyroscope sensor 114. The DSP 111 is used for audio processing, such as encoding and decoding audio in different formats, sound processing, audio playback, and recording. The SPI 112 consists of a master module and one or more slave modules, and the master module selects one of the slave modules for synchronous communication to complete data exchange. The gyroscope sensor 114 is used to detect the user's head posture.

[0040] The codec 120 may include a traditional digital signal processor (TDSP) 121, a fast digital signal processor (FDSP) 122, a serial peripheral interface (SPI) 123, an integrated circuit built-in audio bus (I2S) 124, and a microphone assembly 125. The TDSP and FDSP jointly implement active noise cancellation. The TDSP can analyze the noise characteristics of ambient audio and determine the corresponding noise scene based on its strong computing power. After determining the filter parameters based on the noise characteristics, the TDSP sends the corresponding filter parameters to the FDSP, which then performs efficient noise filtering. The microphone assembly 125 may include a feedforward microphone located outside the earpiece for acquiring ambient audio. The microphone assembly 125 may also include a feedback microphone located inside the earpiece for acquiring audio at the ear. Optionally, the microphone assembly 125 may be a voice microphone.

[0041] In addition, those skilled in the art will understand that the structure of the earphone 100 shown in the above figures does not constitute a limitation on the earphone. The earphone may include more or fewer components than shown, or combine certain components, or have different component arrangements. For example, the earphone 100 may also include a sound-generating unit, a speaker, sensors (such as an accelerometer, angular velocity sensor, light sensor, etc.), audio circuitry, a Wi-Fi (Wireless Fidelity) module, a power supply, a Bluetooth module, and other components, which will not be described in detail here.

[0042] Please refer to Figure 2This document illustrates a flowchart of an audio signal processing method provided in an exemplary embodiment of this application. The embodiments of this application use this method for... Figure 1 Taking the earphone device shown as an example, the method may include the following steps.

[0043] Step 201: The first processor determines the target active noise cancellation mode and generates filter parameters based on the noise characteristics of the environment in which the headphones are located.

[0044] The first processor can be a TDSP with strong computing power. In one possible implementation, the headphones acquire ambient audio through a feedforward microphone. The first processor then performs noise analysis on the ambient audio to determine the noise characteristics of the user's current environment. For ambient audio with different noise characteristics, the user's noise reduction needs will also differ. Therefore, for different noise environments, the first processor determines which active noise cancellation mode to apply; that is, the first processor determines the target active noise cancellation mode based on the noise characteristics.

[0045] Indicative, such as Figure 3 As shown, the first processor performs real-time noise analysis on the ambient audio and then selects a suitable active noise reduction mode as the target active noise reduction mode from the standard noise reduction mode, adaptive noise reduction mode, and directional noise reduction mode based on the noise characteristics obtained from the analysis.

[0046] Furthermore, in different active noise cancellation modes, the first processor uses different methods to determine the filter parameters, and the filter parameters in different active noise cancellation modes have correspondingly different characteristics. In this embodiment, the first processor determines the target active noise cancellation mode based on noise characteristics, that is, determines the filter parameters that match the ambient audio.

[0047] In another possible implementation, when the user's noise environment changes, the first processor can intelligently switch to a target active noise cancellation mode for active noise cancellation processing based on the noise characteristics obtained from noise analysis. For example, when the user is in an office environment, the first processor determines the target active noise cancellation mode as the appropriate mode to cancel out all noise based on the singular noise characteristics of the office environment. When the user finishes work and enters the subway, the ambient audio in the subway is noisy and contains specific scene noises, such as arrival announcements. Based on the change in noise characteristics, the first processor can intelligently switch the target active noise cancellation mode to a noise cancellation mode with specific sound filtering effects.

[0048] Optionally, users can manually adjust the active noise cancellation mode of the application. For example, users can configure it through an app (application) that has active noise cancellation adjustment function, or adjust the active noise cancellation mode through predefined gestures, such as tapping.

[0049] Step 202: The second processor performs active noise reduction processing based on the filter parameters provided by the first processor.

[0050] The second processor can be a fast signal processing FDSP. In one possible implementation, the first processor sends filter parameters to the second processor, which, based on these filter parameters, performs phase inversion processing on the ambient audio according to fixed logic, thereby obtaining an inverted audio with the same amplitude but opposite phase as the ambient audio, thus achieving active noise reduction.

[0051] In summary, in this embodiment, the headphones implement active noise cancellation through dual processors. The first processor determines the noise characteristics of the environment, identifies a target active noise cancellation mode that meets the noise cancellation requirements from multiple active noise cancellation modes, and generates corresponding filter parameters. Upon receiving the filter parameters, the second processor performs active noise cancellation processing based on these parameters. This solution analyzes the noise characteristics of the ambient audio and intelligently determines the target noise cancellation mode that meets the user's needs from multiple active noise cancellation modes suitable for different noise environments. Compared to related technologies that only use one active noise cancellation mode for noise cancellation processing, this improves the quality of active noise cancellation for environmental noise in different scenarios.

[0052] In this embodiment, the headphone device utilizes dual processors for active noise cancellation. Please refer to... Figure 4 The document illustrates a flowchart of an active noise reduction method. Embodiments of this application use this method for... Figure 1 Taking the earphone device shown as an example, the method may include the following steps.

[0053] Step 401: The first processor preprocesses the ambient audio to obtain the spectral information of the ambient audio.

[0054] When the feedforward microphone captures ambient audio, it is converted from digital to analog and then transmitted as PCM (Pulse Code Modulation) data segments. The noise engine in the first processor then performs noise analysis on the ambient audio.

[0055] In one possible implementation, such as Figure 5As shown, when an ambient audio PCM data segment transmitted by the feedforward microphone is acquired, the first processor performs real-time preprocessing on the PCM data segment and stores the resulting spectral information segment in a buffer. During preprocessing, based on the waveform of the ambient audio in the time domain expressed by the PCM data segment, the first processor obtains the ambient audio expressed as a combination of multiple sine waves through Fourier transform and other methods, that is, obtains the frequency domain expression of the ambient audio, and thus obtains the spectral information of the corresponding PCM data segment. The spectral information characterizes the acoustic features of the ambient audio, such as frequency and amplitude.

[0056] Step 402: The first processor extracts noise parameters from the spectrum information to characterize the noise properties.

[0057] In one possible implementation, such as Figure 5 As shown, the first processor extracts noise-related vector model parameters based on spectral information using an algorithm analysis library (Algo Lib). These vector model parameters include acoustic features such as frequency and amplitude that characterize noise properties. Optionally, the first processor can also analyze the spectral information using a neural network (NN) model to obtain noise parameters.

[0058] Step 403: The first processor determines the target active noise reduction mode from at least two active noise reduction modes based on noise parameters.

[0059] The noise parameters mentioned above characterize the noise characteristics of the ambient audio, that is, the environmental characteristics of the noise environment in which the user is located. For different noise environments, users have different requirements for noise reduction effects. Therefore, the first processor can determine the user's noise reduction requirements based on the noise characteristics, that is, determine the corresponding target active noise reduction mode.

[0060] In one possible implementation, such as Figure 5 As shown, the first processor compares the obtained noise parameters with the vector models in the Data Lib, and determines the one that conforms to the characteristics of the ambient audio noise among at least two active noise reduction modes as the target active noise reduction mode.

[0061] Specifically, the first processor matches the noise parameters with the noise models corresponding to different active noise reduction modes to determine the target active noise reduction mode. The matching degree between the noise parameters and the noise model corresponding to the target active noise reduction mode is higher than the matching degree between the noise parameters and the noise models corresponding to other active noise reduction modes.

[0062] In one possible implementation, the active noise cancellation mode may include at least two of the following: standard noise cancellation mode, adaptive noise cancellation mode, and directional noise cancellation mode. The standard noise cancellation mode can achieve all-around noise cancellation, that is, eliminate as much noise as possible from the environment. Based on this characteristic, the standard noise cancellation mode has a higher matching degree for situations where the noise parameters are simple and there are no special frequency spectra. For example, when a user uses the active noise cancellation function in an office, the ambient noise in the office is mostly stable and disordered sounds such as white noise, and the user does not have a need to obtain special ambient noise. Active noise cancellation can be achieved by eliminating all ambient audio while ensuring the user's needs are met.

[0063] Correspondingly, adaptive noise cancellation mode can determine the current noise scene based on noise parameters of the ambient audio, and further determine the ambient audio that the user needs to hear in that scene. Then, while eliminating some noise, it adaptively retains some specific noise. For situations where noise parameters are inconsistent and specific frequencies appear irregularly, adaptive noise cancellation mode has a higher matching degree. For example, when a user uses active noise cancellation in the subway, the ambient noise is noisy, disorderly, and changes rapidly, with random arrival announcements that the user needs. In adaptive noise cancellation mode, the headphone device can filter out background noise while retaining specific ambient noise such as arrival announcements based on the scene to meet the user's needs. In other words, the user can obtain specific types of noise through adaptive noise cancellation mode.

[0064] In directional noise cancellation mode, the headphone device can determine the direction of each sound source based on information such as sound pressure difference and time difference received by the microphone array. When noise from a specific direction is detected to persist for a certain period of time, the directional noise cancellation mode can determine that this noise is a specific noise that the user is likely to need, and then selectively retain this specific noise during the noise cancellation process. For situations where a specific noise parameter is persistently present, the directional noise cancellation mode has a higher matching accuracy. For example, when a user uses active noise cancellation in a factory environment, and the conversation of colleagues comes from a fixed direction and persists for a certain period of time, the directional noise cancellation mode can retain this noise based on the determined direction of the conversation source. In other words, the user can obtain specific noise with a specific direction through the directional noise cancellation mode.

[0065] Step 404: The first processor generates first filter parameters based on the working scenario of the headphones; and / or generates second filter parameters based on the ambient audio from at least one sound source direction of the sound source received by the headphones.

[0066] Based on different noise characteristics, that is, for different ambient audio environments, the headphones employ different active noise cancellation modes. During the noise cancellation process, each active noise cancellation mode achieves its corresponding noise reduction effect by using different filter parameters. Noise filtering can be performed using any of the following methods:

[0067] 1. Generate the first filter parameters based on the working scenario of the headphones.

[0068] 1.1 When the target active noise cancellation mode is adaptive noise cancellation mode, the working scenario of the headphones is determined based on the ambient audio; the first filter parameters are generated according to the working scenario. Different working scenarios correspond to different filter parameters.

[0069] The structure of the adaptive noise reduction mode is as follows Figure 8 As shown, the headphones first acquire ambient noise x(t) through a feedforward microphone to obtain ambient audio x(n). Then, the first processor analyzes the ambient audio for noise and determines its noise characteristics. Given the noisy environment containing randomly occurring specific noise, the first processor employs an adaptive noise reduction mode for active noise cancellation. Based on the adaptive noise reduction mode, specific noise that needs to be retained is determined according to the noise environment. The adaptive noise reduction mode uses a scene recognition unit to determine a relatively accurate working scene, such as an outdoor scene, a subway scene, or a coffee shop scene. Based on the working scene, the first processor obtains the corresponding filter parameters. The first filter parameters differ depending on the frequency of the noise that needs to be retained in different scenes.

[0070] If the headphones are equipped with a feedback microphone, such as Figure 7 As shown, the feedback microphone collects the feedback audio e(n), and the feedback filter performs feedback filtering based on the feedback audio. In the adaptive noise reduction mode, the first processor obtains the second target filtering parameters applied to the feedback filter and sends them to the feedback filter in the second processor.

[0071] Optionally, since ambient audio varies significantly in noisy environments, using variable filter parameters can improve filtering performance. Therefore, adaptive noise reduction modes can be combined with adaptive algorithms to improve active noise reduction in noisy environments. For example, an adaptive noise reduction mode can employ the Least Mean Square (LMS) algorithm, periodically adjusting filter parameters based on feedback audio collected by the feedback microphone to reduce the error between the target noise and the determined noise. In the active noise reduction process, the LMS algorithm can be implemented through the following calculation process:

[0072] y(k)=w T (k)x(k)

[0073] e(k) = d(k) - y(k)

[0074] w(k+1)=w(k)+μe(k)x(k)

[0075] Where y(k) is the desired output, w T (k) represents the filter coefficient vector in the k-th iteration, x(n) is the input signal, which is the vector composed of N values ​​collected in the most recent time period, and e(k) represents the noise error between the target noise d(k) and the desired output y(k). The least mean square algorithm is a special type of gradient descent calculation, where μ is the step size of gradient descent, and w(k+1) is the filter coefficient vector in the (k+1)-th iteration. This calculation process is carried out through iterative loops to make y(n) as close as possible to d(n), that is, to minimize the error e(n).

[0076] 2. Generate second filter parameters based on the ambient audio from at least one direction of the sound source received by the headphones.

[0077] 2.1 When the target active noise cancellation mode is the standard noise cancellation mode, the second filter parameters are generated based on the ambient audio from each sound source direction received by the headphones.

[0078] The structure of the standard noise reduction mode is as follows: Figure 6 As shown, the headphones acquire ambient audio x(n) by collecting external noise x(t) through a feedforward microphone. During the transmission of external noise from the headphones to the ear, i.e., through the headphone shell and cavity in the main sound path, the sound attenuates during transmission. The attenuated noise audio obtained at the ear is x'(t), which can be represented by the main sound transfer function P(n). The first processor performs noise analysis on the ambient audio x(n) and determines the noise parameters characterizing its noise properties. Based on the single, disordered noise reduction parameters, the first processor determines the standard noise reduction mode as the target active noise reduction mode. In this mode, because the ambient audio is disordered and stable, the first processor can determine the second filter parameters based on the omnidirectional ambient audio received from various sound source directions. In this case, the first processor can achieve a better noise reduction effect by using predefined filter parameters. That is, the first processor obtains the fixed filter parameters corresponding to the standard noise reduction mode as the second filter parameters, such as white noise filter parameters, and sends them to the second processor.

[0079] In one possible implementation, the headphones may also include a feedback microphone for capturing feedback audio. In this case, the first processor, based on the target active noise cancellation mode, sends a first target filtering parameter and a second target filtering parameter to the second processor. The first target filtering parameter is the parameter used by the feedforward filter in the second processor for audio filtering processing, and the second target filtering parameter is the parameter used by the feedback filter in the second processor for audio filtering processing. Both the first and second target filtering parameters belong to the aforementioned filter parameters. Figure 7 As shown, in the case of a feedback microphone, the feedback microphone collects the feedback audio e(n) at the human ear. The first processor performs noise analysis on the feedback audio to determine the second target filtering parameters, and sends them to the feedback filter in the second processor for feedback filtering.

[0080] 2.2 When the target active noise cancellation mode is directional noise cancellation mode, the direction of the target sound source is determined based on the ambient audio received by the headphones; the second filter parameters are generated based on the ambient audio from directions other than the target sound source direction.

[0081] The structure of the directional noise reduction mode is as follows Figure 9 As shown, compared to the standard noise reduction mode, which only uses a feedforward filter or combines it with a feedback filter for noise reduction, the directional noise reduction mode introduces a directional noise reduction unit to help determine the specific noise that needs to be retained. Based on the first processor's use of the directional noise reduction mode when there is a specific direction of continuous, special noise in the ambient audio, and the corresponding noise environment has a certain time-varying nature, this noise reduction mode can combine adaptive algorithms such as LMS to determine the second filter parameters and optimize the noise reduction effect. Specifically, the first processor can determine the direction of the continuous noise with characteristic audio information as the target sound source direction based on the ambient audio. Optionally, in the directional noise reduction mode, the first processor can obtain corresponding predefined filter parameters as the second filter parameters based on this mode. Optionally, the second filter parameters can be further optimized based on the ambient audio in directions other than the target sound source direction.

[0082] Step 405: The second processor configures the filter based on the filter parameters and performs filtering processing on the ambient audio through the filter.

[0083] The filter parameters can be either first filter parameters or second filter parameters. The second processor can be an FDSP. In one possible implementation, the FDSP can include multiple filters, and if the headphones are equipped with a feedback filter, the multiple filters can simultaneously filter the ambient audio captured by the feedforward microphone and the feedback audio captured by the feedback microphone.

[0084] The process of filtering based on filter parameters by the second processor will be explained using the adaptive noise cancellation mode of headphones as an example. Figure 8 As shown, the second processor determines the filter coefficient vector w (expressed as W1(z) in the frequency domain) suitable for the feedforward filter based on the filter parameters sent by the first processor. Then, the feedforward filter determines the desired output y(n) based on the filter coefficient vector w and the input signal x(n), i.e., the ambient audio.

[0085] Step 406: The second processor generates inverted audio based on the filtered ambient audio. The inverted audio has the opposite phase to the filtered ambient audio.

[0086] In one possible implementation, such as Figure 10 As shown, after obtaining the PCM (audio data) of the ambient audio, the second processor first preprocesses it and filters it based on the filter parameters to obtain an inverted audio with the same amplitude but opposite phase as the ambient audio. Then, the inverted audio cancels out the ambient audio to achieve active noise reduction.

[0087] When the headphones have a feedback microphone, taking adaptive noise cancellation mode as an example, such as... Figure 8 As shown, both the feedforward filter and the feedback filter generate inverted audio accordingly. During the adaptive noise reduction process, the more accurate inverted audio can be determined by combining the two digital audio signals through accumulation operations.

[0088] Step 407: The second processor transmits inverted audio to the speaker so that the speaker plays inverted audio.

[0089] Upon determining that the audio is inverted, the second processor transmits the inverted audio to the speaker. In one possible implementation, taking an adaptive noise reduction mode as an example, such as... Figure 8 As shown, the inverted audio signal is converted from a digital signal to an analog signal via a digital-to-analog converter and transmitted to an AMP (Amplifier), where the AMP amplifies the analog signal. The inverted audio signal y(n) is then played by a speaker. The fundamental sound travels from the speaker to the ear and undergoes a secondary sound filter, resulting in some attenuation. This attenuation can be represented by the secondary sound transfer function S(z). The attenuated audio signal is y'(t). When the sound played by the speaker and ambient noise are simultaneously transmitted to the ear, the audio signals with equal amplitude but opposite phase in x'(t) and y'(t) cancel each other out through interference, thus achieving active noise reduction.

[0090] In summary, this application embodiment processes noise using a dual-processor architecture. The first processor can be a TDSP capable of flexibly controlling its operating program. The first processor analyzes ambient audio noise, determines noise parameters, and intelligently selects an active noise cancellation mode that better matches user needs based on these parameters, achieving intelligent noise reduction without needing to wear the speaker. Furthermore, the first processor determines filter parameters suitable for the user's noise reduction requirements and sends these parameters to the second processor. The second processor can be an FDSP with fixed operational logic and high processing speed. Based on the filter parameters, the second processor obtains the inverted audio of the ambient audio through the filter. The speaker then plays the inverted audio to cancel interference, completing active noise cancellation tailored to user needs. This application intelligently determines the active noise cancellation mode through noise analysis, improving the quality of active noise cancellation for ambient noise in different scenarios.

[0091] In a specific scenario, there exists a particular type of noise in the environment with a relatively fixed sound source location and a long duration. This ambient sound is highly likely to be the specific noise that the user needs to hear. In this case, the headphones in this solution employ a directional noise cancellation mode based on the aforementioned noise characteristics to assist the user in hearing the sound. Please refer to [link / reference]. Figure 11 The diagram illustrates a flowchart of a directional noise reduction mode provided in an exemplary embodiment of this application.

[0092] In this embodiment, the earphone is equipped with multiple feedforward microphones, each used to receive ambient audio from different directions. Optionally, the feedforward microphones can be voice microphones, and multiple voice microphones form a microphone array. The topology of the multiple microphones in this array can be arranged in a straight line or in a honeycomb pattern, etc., and this solution does not limit this.

[0093] Step 1101: The first processor uses a beamformer to perform beamforming processing based on ambient audio to obtain directional audio. The beamforming processing is used to generate an audio compensation signal in the direction of the target sound source to cancel the ambient audio in at least one sound source direction outside the direction of the target sound source.

[0094] like Figure 9 As shown, the directional noise reduction mode includes a directional unit 901, which comprises multiple microphones for acquiring multiple ambient audio signals and a beamformer for beamforming the multiple ambient audio signals. In one possible implementation, as... Figure 12As shown, the left and right earphones each have two codecs, namely the left codec 1210 and the right codec 1220. In addition to TDSPs 1211 and 1221 and FDSPs 1212 and 1222 respectively, the decoders each have three microphones for capturing ambient audio, forming microphone arrays 1213 and 1223. Furthermore, as... Figure 13 As shown, when both earbuds are worn, they monitor ambient noise within a 180° range on each side via their respective microphone arrays. It should be noted that this solution does not limit the method of information exchange between the left and right earbuds, and correspondingly, it does not limit the number of microphones used for ambient audio acquisition in either earbud.

[0095] For multiple ambient audio streams, the first processors in the left and right earbuds respectively use beamforming algorithms to determine the location of the corresponding sound source based on the time difference and sound pressure difference of each ambient audio stream arriving at the microphone array. The left and right earbuds can exchange information via Bluetooth, forming a beamforming array. By combining the sound source information from both earbuds, the earbuds initially determine specific noises that the user is likely to encounter through beamforming. Combining this with other noise in the ambient audio, the first processor merges the multiple ambient audio streams and determines the directional audio, enhancing the ambient audio in the direction of the target sound source. Correspondingly, the beamformer generates an audio compensation signal through beamforming processing. A subsequent second processor can then use this audio compensation signal to reduce noise in the ambient audio from other directions, thus filtering out background noise while retaining and enhancing the ambient noise that the user needs to acquire.

[0096] Optionally, users can use the terminal application to set the type of environmental noise that needs to be identified as special noise. For example, users can set to select specific human voices, specific types of sounds, or perform adaptive selection. Correspondingly, users can manually collect specific human voices or specific types of sounds and store the corresponding sound spectrum information.

[0097] To illustrate, when someone is speaking continuously to the user's left, the sound pressure detected by the microphone in the microphone array closer to the direction of the audio source is greater than the sound pressure detected by the microphone in the direction of the original sound source. Combined with the fact that the sound lasts for a certain duration, or meets predefined special audio acoustic characteristics (pre-collected and stored spectrum information), the directional unit can determine that the sound from that direction is special noise. Furthermore, during the merging of multiple ambient audio streams, the first processor determines that the noise from the aforementioned sound source direction is amplified, while the noise from the other directions is suppressed.

[0098] Step 1102: The first processor sends directional audio to the second processor.

[0099] When a directional audio signal is determined, the first processor sends the directional audio signal to the second processor so that the second processor can filter the directional audio signal as an input signal.

[0100] Step 1103: The second processor generates target audio based on the inverted audio and directional audio.

[0101] Based on the aforementioned directional audio, the second processor uses filters to determine the specific noise audio from a specific direction, as well as the remaining noise audio, and generates audio signals that constructively interfere with the specific noise audio to enhance it, and generates audio signals that destructively interfere with the remaining noise audio to cancel it out, thus achieving active noise reduction while preserving the specific noise.

[0102] like Figure 9 As shown, selective noise reduction is achieved by merging ANC and beamforming based on the directional noise reduction mode. The second processor obtains the inverted audio through the ANC process and accumulates the audio obtained after directional audio filtering with the inverted audio to obtain the target audio.

[0103] Step 1104: The second processor transmits the target audio to the speaker so that the speaker plays the target audio.

[0104] This step is the same as step 407, and will not be repeated here.

[0105] In one possible implementation, the earphones are equipped with sensors. Upon determining a specific orientation, a first processor determines the earphones' attitude change parameters based on acquired sensor data, which includes at least the earphones' vector rotational angular acceleration.

[0106] The aforementioned sensor can be a gyroscope sensor. Based on beamforming, the user can acquire ambient noise from a specific direction. When the user's head rotates, the headphones can determine the direction and angle of the user's head rotation using gyroscope sensor data. Based on this data, the headphones can determine the user's response to the acquired ambient noise. The first processor can then further refine the localization of specific noises based on the headphones' attitude change parameters, improving the noise reduction accuracy of the directional noise cancellation mode. The attitude change parameters include at least the rotation time, rotation amplitude, and direction.

[0107] In one possible implementation, when the gyroscope-assisted positioning function is enabled, the earphone acquires gyroscope sensor data in real time based on a sampling frequency f. This gyroscope sensor data includes vector angular acceleration values ​​b in the X, Y, and Z axes. t (x, y, z). Set the key-value pair k. t =(r t ,bt ), where r t b is the sampling time. t The vector angular acceleration values ​​are obtained based on a set time window T seconds, with a number of N values. T The data is stored in a circular queue in the data structure K[N] T In ], where N T =T×f and have In directional noise cancellation mode, when the first processor initially detects special noise, it acquires gyroscope sensor data based on the headphones. From the moment the special noise is received, the first processor searches for key-value pairs and performs tracking and calculation.

[0108] To illustrate, when a user uses active noise cancellation in a factory environment, a coworker speaks continuously from a 30° angle to the user's left front. Through initial beamforming, the first processor, while merging multiple ambient audio streams, determines that the 45° range to the user's left front is the source of specific noise, amplifying ambient noise from this range and reducing ambient noise from other directions. When the user hears the conversation and turns their head towards the coworker, gyroscope sensor data characterizes the direction and magnitude of the user's head turn, and can further determine that the direction of the coworker's voice audio is the user's facing direction, i.e., the 30° angle to the user's left front.

[0109] Furthermore, the first processor corrects the specific direction based on the attitude change parameters. And, if the attitude change parameters exceed a threshold, the first processor corrects the specific direction based on the attitude change parameters.

[0110] If a change in gyroscope sensor data is detected after receiving a specific noise, and the attitude change parameter is greater than a threshold, the earphone can determine that the user has reacted to the specific noise. If the initially determined direction of the sound source has a certain error, and the gyroscope sensor data reflects that the user is turning towards the vicinity of the initially determined direction of the specific noise sound source, the first processor combines the gyroscope sensor data to correct the current direction of the user to a specific direction.

[0111] Furthermore, the first processor performs beamforming on the multiple ambient audio streams based on the corrected specific direction to obtain directional audio. During beamforming, the first processor further enhances the ambient noise from the corrected specific direction, while correspondingly filtering out ambient noise from other directions, thus improving the directional noise reduction effect.

[0112] In summary, this solution uses beamforming to detect ambient noise in real time and preserves specific ambient noise during active noise cancellation. This ensures that users can still access ambient noise that they need, such as voices and alarms, even when using active noise cancellation. Building upon beamforming for directional noise cancellation, this solution combines beamforming with sensors such as gyroscopes to detect user head movements and uses head rotation information to determine the user's response to specific noises. This further improves the accuracy of locating specific noises, enhances the effectiveness of active noise cancellation, and improves its overall quality.

[0113] Please refer to Figure 14 The diagram illustrates a structural block diagram of an active noise cancellation device provided in an exemplary embodiment of this application, the device comprising:

[0114] The first processor 1401 is used to determine the target active noise cancellation mode and generate filter parameters based on the noise characteristics of the environment in which the headphones are located.

[0115] The second processor 1402 is used to perform active noise reduction processing based on the filter parameters provided by the first processor.

[0116] Optionally, the first processor 1401 is further configured to:

[0117] The ambient audio is preprocessed to obtain the spectral information of the ambient audio.

[0118] Noise parameters characterizing the noise properties are extracted from the spectrum information;

[0119] Based on the noise parameters, the target active noise reduction mode is determined from at least two active noise reduction modes.

[0120] Optionally, the first processor 1401 is further configured to:

[0121] The noise parameters are matched with the noise models corresponding to different active noise reduction modes to determine the target active noise reduction mode.

[0122] Optionally, the first processor 1401 is further configured to:

[0123] Generate first filter parameters based on the operating scenario of the headphones; and / or

[0124] The second filter parameters are generated based on the ambient audio from at least one sound source direction received by the headphones.

[0125] Optionally, when the active noise cancellation mode includes a standard noise cancellation mode, an adaptive noise cancellation mode, and a directional noise cancellation mode, the first processor 1401 is further configured to:

[0126] When the target active noise cancellation mode is the adaptive noise cancellation mode, the working scenario of the headphones is determined based on the ambient audio; and the first filter parameters are generated according to the working scenario.

[0127] When the target active noise cancellation mode is the standard noise cancellation mode, the second filter parameters are generated based on the ambient audio from each sound source direction received by the headphones.

[0128] When the target active noise cancellation mode is directional noise cancellation mode, the direction of the target sound source is determined based on the ambient audio received by the headphones; and the second filter parameters are generated based on the ambient audio in the direction of the target sound source.

[0129] Optionally, the second processor 1402 is further configured to:

[0130] Configure the filter based on the filter parameters, and use the filter to filter ambient audio.

[0131] An inverted audio is generated based on the filtered ambient audio, and the inverted audio is out of phase with the filtered ambient audio.

[0132] The inverted audio is transmitted to the speaker so that the speaker plays the inverted audio.

[0133] Optionally, when the headphones are equipped with multiple feedforward microphones, and different feedforward microphones are used to receive ambient audio from different directions, the first processor 1401 is further configured to:

[0134] The first processor uses a beamformer to perform beamforming processing based on the ambient audio to obtain directional audio. The beamforming processing is used to generate an audio compensation signal in the direction of the target sound source to cancel the ambient audio in at least one sound source direction outside the direction of the target sound source.

[0135] The first processor sends the directional audio to the second processor;

[0136] The second processor 1402 is also used for:

[0137] The second processor generates the target audio based on the inverted audio and the directional audio;

[0138] The second processor transmits the target audio to the speaker so that the speaker plays the target audio.

[0139] Optionally, if a sensor is provided in the earphone for determining the motion state of the earphone, the first processor 1401 is further configured to:

[0140] Once the direction of the target sound source is determined, the first processor determines the posture change parameters of the headphones based on the acquired sensor data, wherein the sensor data includes at least the vector rotational angular acceleration of the headphones;

[0141] The first processor corrects the direction of the target sound source based on the attitude change parameters;

[0142] The first processor performs beamforming processing on multiple ambient audio streams based on the corrected target sound source direction to obtain the directional audio.

[0143] Optionally, the first processor 1401 is further configured to:

[0144] If the attitude change parameter is greater than a threshold, the direction of the target sound source is corrected based on the attitude change parameter.

[0145] Optionally, if the headphones are further equipped with a feedback microphone for acquiring feedback audio, the first processor 1401 is further configured to:

[0146] Based on the target active noise reduction mode, a first target filtering parameter and a second target filtering parameter are sent to the second processor. The first target filtering parameter is the parameter used by the feedforward filter in the second processor for audio filtering processing, and the second target filtering parameter is the parameter used by the feedback filter in the second processor for audio filtering processing.

[0147] Optionally, the first processor 1401 and the second processor 1402 are also used to perform noise reduction processing in parallel.

[0148] In summary, this application embodiment implements noise reduction functionality through dual processors. The first processor, based on the noise characteristics of the ambient audio collected by the feedforward microphone, determines a more suitable target active noise reduction mode and generates filter parameters among multiple active noise reduction modes. Upon receiving the filter parameters, the second processor performs active noise reduction processing based on those parameters. This solution analyzes the noise characteristics of the ambient audio and intelligently determines the target noise reduction mode that meets the user's needs from multiple active noise reduction modes suitable for different noise environments. Compared to related technologies that only use one fixed active noise reduction mode for noise reduction processing, this improves the quality of active noise reduction for ambient noise in different scenarios.

[0149] This application embodiment also provides a chip, which includes a first processor and a second processor. The first processor is configured to: determine a target active noise cancellation mode based on the noise characteristics of the environment in which the headphones are located and generate filter parameters; the second processor is configured to: perform the active noise cancellation processing described in the above embodiment based on the filter parameters provided by the first processor.

[0150] This application also provides a computer-readable storage medium storing at least one program that is executed by a processor to implement the active noise reduction method as described in the above embodiments.

[0151] This application provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the active noise reduction method provided in the above embodiments.

[0152] Those skilled in the art will recognize that the functions described in the embodiments of this application in one or more of the above examples can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transfer of a computer program from one place to another. Storage media can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0153] The above description is merely an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. An active noise reduction method, characterized in that, The method is applied to headphones, which include a first processor and a second processor. The headphones are equipped with multiple feedforward microphones, each used to receive ambient audio from different directions. The headphones also include a sensor for determining the motion state of the headphones. The method includes: The first processor preprocesses the ambient audio to obtain the spectral information of the ambient audio; extracts noise parameters from the spectral information to characterize the noise characteristics of the environment in which the headphones are located; and determines a target active noise cancellation mode from at least two active noise cancellation modes based on the noise parameters, wherein the active noise cancellation mode includes a standard noise cancellation mode, an adaptive noise cancellation mode, and a directional noise cancellation mode. When the target active noise cancellation mode is the adaptive noise cancellation mode, the first processor determines the working scenario of the headphones based on the ambient audio; and generates filter parameters according to the working scenario. When the target active noise cancellation mode is the standard noise cancellation mode, the first processor generates filter parameters based on the ambient audio received by the headphones from each sound source direction. When the target active noise cancellation mode is the directional noise cancellation mode, the first processor determines the target sound source direction based on the ambient audio received by the headphones; and generates filter parameters based on the ambient audio from directions other than the target sound source direction. When the target active noise cancellation mode is directional noise cancellation mode and the direction of the target sound source is determined, the first processor determines the posture change parameters of the earphone based on the acquired sensor data, wherein the sensor data includes at least the vector rotation angular acceleration of the earphone; the first processor corrects the direction of the target sound source based on the posture change parameters; the first processor performs beamforming processing on multiple ambient audio streams based on the corrected direction of the target sound source to obtain directional audio, wherein the beamforming processing is used to generate an audio compensation signal in the direction of the target sound source to cancel the ambient audio in at least one sound source direction other than the direction of the target sound source, and the directional audio is used to enhance the ambient audio in the direction of the target sound source; the first processor sends the directional audio to the second processor; The second processor configures a filter based on filter parameters and performs filtering processing on the ambient audio through the filter; it generates inverted audio based on the filtered ambient audio, the inverted audio being out of phase with the filtered ambient audio. When the target active noise cancellation mode is the adaptive noise cancellation mode or the standard noise cancellation mode, the second processor transmits the inverted audio to the speaker so that the speaker plays the inverted audio; when the target active noise cancellation mode is the directional noise cancellation mode, the second processor performs an accumulation operation on the inverted audio and the directional audio to generate the target audio; the second processor transmits the target audio to the speaker so that the speaker plays the target audio.

2. The method according to claim 1, characterized in that, The step of determining the target active noise reduction mode from at least two active noise reduction modes based on the noise parameters includes: The noise parameters are matched with the noise models corresponding to different active noise reduction modes to determine the target active noise reduction mode.

3. The method according to claim 1, characterized in that, The first processor corrects the direction of the target sound source based on the attitude change parameters, including: If the attitude change parameter is greater than a threshold, the direction of the target sound source is corrected based on the attitude change parameter.

4. The method according to claim 1, characterized in that, The headphones are also equipped with a feedback microphone, which is used to collect feedback audio. The method further includes: Based on the target active noise reduction mode, the first processor sends a first target filtering parameter and a second target filtering parameter to the second processor. The first target filtering parameter is the parameter used by the feedforward filter in the second processor for audio filtering processing, and the first target filtering parameter is the parameter used by the feedback filter in the second processor for audio filtering processing.

5. The method according to claim 1, characterized in that, The first processor and the second processor perform noise reduction processing in parallel.

6. An active noise reduction device, characterized in that, The device is used for headphones, the headphones including a first processor and a second processor, the headphones being provided with multiple feedforward microphones, different feedforward microphones being used to receive ambient audio from different directions, and the headphones being provided with sensors for determining the motion state of the headphones, the device comprising: A first processor is configured to preprocess the ambient audio to obtain the spectral information of the ambient audio; extract noise parameters from the spectral information to characterize the noise characteristics of the environment in which the headphones are located; and determine a target active noise cancellation mode from at least two active noise cancellation modes based on the noise parameters, wherein the active noise cancellation mode includes a standard noise cancellation mode, an adaptive noise cancellation mode, and a directional noise cancellation mode. A first processor is configured to: determine the operating scenario of the headphones based on the ambient audio when the target active noise cancellation mode is the adaptive noise cancellation mode; generate filter parameters according to the operating scenario; generate filter parameters based on the ambient audio received by the headphones from each sound source direction when the target active noise cancellation mode is the standard noise cancellation mode; determine the target sound source direction based on the ambient audio received by the headphones when the target active noise cancellation mode is the directional noise cancellation mode; and generate filter parameters based on the ambient audio from directions other than the target sound source direction. A first processor is configured to, when the target active noise cancellation mode is directional noise cancellation mode and the direction of the target sound source is determined, determine the posture change parameters of the earphone based on acquired sensor data, wherein the sensor data includes at least the vector rotation angular acceleration of the earphone; correct the target sound source direction based on the posture change parameters; perform beamforming processing on multiple ambient audio streams based on the corrected target sound source direction to obtain directional audio, wherein the beamforming processing is used to generate an audio compensation signal in the direction of the target sound source to cancel the ambient audio in at least one sound source direction other than the target sound source direction, wherein the directional audio is used to enhance the ambient audio in the direction of the target sound source; and send the directional audio to a second processor. The second processor is configured to configure a filter based on filter parameters and perform filtering processing on the ambient audio through the filter; and generate inverted audio based on the filtered ambient audio, wherein the inverted audio is out of phase with the filtered ambient audio. The second processor is configured to, when the target active noise cancellation mode is the adaptive noise cancellation mode or the standard noise cancellation mode, transmit the inverted audio to the speaker so that the speaker plays the inverted audio; when the target active noise cancellation mode is the directional noise cancellation mode, perform an accumulation operation on the inverted audio and the directional audio to generate target audio; and transmit the target audio to the speaker so that the speaker plays the target audio.

7. A chip, characterized in that, The chip is used in headphones, and the chip includes a first processor and a second processor, wherein the first processor is configured to: The ambient audio is preprocessed to obtain the spectral information of the ambient audio. The earphone is equipped with multiple feedforward microphones, and different feedforward microphones are used to receive the ambient audio from different directions. The earphone is also equipped with a sensor, which is used to determine the motion state of the earphone. Noise parameters that characterize the noise characteristics of the environment in which the earphone is located are extracted from the spectral information. Based on the noise parameters, a target active noise reduction mode is determined from at least two active noise reduction modes, wherein the active noise reduction mode includes a standard noise reduction mode, an adaptive noise reduction mode, and a directional noise reduction mode; When the target active noise cancellation mode is the adaptive noise cancellation mode, the working scenario of the headphones is determined based on the ambient audio; filter parameters are generated according to the working scenario; when the target active noise cancellation mode is the standard noise cancellation mode, filter parameters are generated according to the ambient audio received by the headphones from each sound source direction; when the target active noise cancellation mode is the directional noise cancellation mode, the target sound source direction is determined based on the ambient audio received by the headphones; filter parameters are generated according to the ambient audio from directions other than the target sound source direction. When the target active noise cancellation mode is directional noise cancellation mode and the direction of the target sound source is determined, the attitude change parameters of the headphones are determined based on the acquired sensor data, the sensor data including at least the vector rotation angular acceleration of the headphones; the direction of the target sound source is corrected based on the attitude change parameters; beamforming processing is performed on multiple ambient audio streams based on the corrected direction of the target sound source to obtain directional audio, the beamforming processing is used to generate an audio compensation signal in the direction of the target sound source to cancel the ambient audio in at least one sound source direction other than the direction of the target sound source, the directional audio is used to enhance the ambient audio in the direction of the target sound source; the directional audio is sent to the second processor; The second processor is configured as follows: A filter is configured based on filter parameters, and the ambient audio is filtered using the filter. An inverted audio is generated based on the filtered ambient audio, and the inverted audio is out of phase with the filtered ambient audio. When the target active noise cancellation mode is the adaptive noise cancellation mode or the standard noise cancellation mode, the inverted audio is transmitted to the speaker so that the speaker plays the inverted audio; when the target active noise cancellation mode is the directional noise cancellation mode, the inverted audio and the directional audio are accumulated to generate the target audio; the target audio is transmitted to the speaker so that the speaker plays the target audio.

8. An earphone, characterized in that, The earphone includes a processor and a memory, the processor including at least a first processor and a second processor, and the memory storing at least one program, which is loaded and executed by the processor to implement the active noise cancellation method as described in any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one program, which is loaded and executed by a processor to implement the active noise reduction method as described in any one of claims 1 to 5.

10. A computer program product, characterized in that, The computer program product includes computer instructions stored in a computer-readable storage medium; a processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the active noise reduction method as described in any one of claims 1 to 5.

Images