Audio processing method, mouth wearing device, electronic device and medium

Abstract

The invention provides an audio processing method, mouth wearing equipment, electronic equipment and a medium, the method is applied to the field of intelligent wearing equipment, and the method comprises the following steps: when an active silencing function of the mouth wearing equipment is in an enabled state, obtaining a first audio signal of a wearing user within a preset duration; audio feature analysis is carried out on the first audio signal to obtain a predicted audio signal, an active silencing device is called to generate a reverse sound wave signal corresponding to the predicted audio signal, and the reverse sound wave signal is an audio signal with the phase opposite to that of the predicted audio signal; and calling the active silencing device to output a reverse sound wave signal. According to the method, active noise elimination and passive noise elimination are combined, so that full-band effective noise reduction is realized, and the voice privacy protection effect is remarkably improved.

Description

Technical Field

[0001] This specification relates to the field of smart wearable devices, and in particular to audio processing methods, mouth-worn devices, electronic devices, and media. Background Technology

[0002] Existing passive noise-canceling masks require multiple layers of materials to ensure maximum sound elimination, resulting in a bulky device that needs to be tightly fitted to the head, affecting aesthetics. Furthermore, due to space limitations, they can only eliminate high-frequency components of speech, while some low-frequency sounds can still propagate through the mask body by penetrating the materials and vibrating the mask's surface, leading to poor noise cancellation. Therefore, there is a need for a smart mask solution that can effectively protect personal privacy. Summary of the Invention

[0003] This specification provides an audio processing method, a mouth-worn device, an electronic device, and a medium, aiming to improve the privacy and security of users when using the wearable device. The technical solution is as follows: In a first aspect, embodiments of this specification provide an audio processing method applied to a processing device. The processing device is applied to a mouth-worn device, which includes a housing, a breathing valve, an external speaker, a sound-absorbing cavity, and an active sound-absorbing device. A recessed area in the housing forms an accommodating space. The breathing valve is disposed on the housing to connect the accommodating space with the ambient space outside the housing. The external speaker is disposed on the housing for collecting audio signals within the accommodating space. The sound-absorbing cavity is disposed within the accommodating space to cover the mouth area of ​​the user. The active sound-absorbing device is embedded within the cavity of the sound-absorbing cavity. The method includes: When the active noise cancellation function of the mouth-worn device is enabled, the first audio signal of the wearer within a preset duration is acquired. The first audio signal is subjected to audio feature analysis to obtain the predicted audio signal. The active noise cancellation device is called to generate the reverse sound wave signal corresponding to the predicted audio signal. The reverse sound wave signal is an audio signal with the opposite phase to the predicted audio signal. The active silencing device is activated to output a reverse acoustic wave signal.

[0004] Secondly, this specification provides an oral wearing device, which includes a processing device, a housing, a breathing valve, an external speaker, a sound-absorbing cavity, and an active sound-absorbing device. A recessed area in the housing forms an accommodating space. The breathing valve is disposed on the housing to connect the accommodating space with the ambient space outside the housing. The external speaker is disposed on the housing to output audio signals from the accommodating space. The sound-absorbing cavity is disposed within the accommodating space to cover the mouth area of ​​the user. The active sound-absorbing device is embedded within the cavity of the sound-absorbing cavity. The processing device is used to acquire the first audio signal of the wearer within a preset duration when the active noise cancellation function of the mouth-worn device is enabled. Audio feature analysis is performed on the first audio signal to obtain the predicted audio signal; The active noise cancellation device is invoked to generate the reverse acoustic wave signal corresponding to the predicted audio signal. The reverse acoustic wave signal is an audio signal with the opposite phase to the predicted audio signal. The active silencing device is activated to output a reverse acoustic wave signal.

[0005] Thirdly, embodiments of this specification provide an electronic device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements any of the methods described above.

[0006] Fourthly, embodiments of this specification provide a computer-readable storage medium on which a computer program is stored, and when executed, the computer program implements any of the methods described above.

[0007] Fifthly, embodiments of this specification provide a computer program product that stores at least one instruction, which, when executed by a processor, implements any of the methods described above.

[0008] In the embodiments described in this specification, by predicting the user's speech in real time and generating a reverse sound wave, the propagation path of the sound wave is directly intervened, achieving efficient noise cancellation without relying on multi-layer material stacking. Especially for low-frequency sound waves, the principle of phase interference is used for cancellation, overcoming the problem of poor noise cancellation in the low-frequency band of traditional passive noise cancellation devices. At the same time, by reducing the use of physical sound-absorbing materials, the thickness and weight of the device are significantly reduced, improving wearing comfort and aesthetics. Furthermore, by combining active and passive noise cancellation, effective noise reduction is achieved across the entire frequency band, significantly improving voice privacy protection. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of a scenario for an audio processing method provided in an embodiment of this specification; Figure 2 This is a schematic diagram of a scenario for an audio processing method provided in an embodiment of this specification; Figure 3 This is a schematic diagram of a scenario for an audio processing method provided in an embodiment of this specification; Figure 4 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification; Figure 5 This is a schematic diagram of a scenario for an audio processing method provided in an embodiment of this specification; Figure 6 This is a schematic diagram of a scenario for an audio processing method provided in an embodiment of this specification; Figure 7 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification; Figure 8 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification; Figure 9 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification; Figure 10 This is a schematic diagram of a scenario for an audio processing method provided in an embodiment of this specification; Figure 11 This is a schematic diagram of the structure of an electronic device provided in the embodiments of this specification.

[0011] Figure descriptions: 10. Outer shell; 11. Breathing valve; 12. External loudspeaker; 13. Reverse loudspeaker; 14. Silencing cavity; 15. Active silencer; 16. First microphone; 17. Second microphone; 18. First rubber sealing ring; 19. Second rubber sealing ring; 20. Elastic rope; 21. Tension adjuster; 22. Sound-absorbing foam; 23. Processing device. Detailed Implementation

[0012] To make the features and advantages of this specification more apparent and understandable, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this specification without creative effort are within the scope of protection of this specification.

[0013] The technical solutions in this specification will now be described clearly and in detail with reference to the accompanying drawings. In the description of the embodiments in this specification, unless otherwise stated, " / " indicates "or," for example, A / B can mean A or B. "And / or" in the text is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of the embodiments in this specification, "multiple" refers to two or more than two.

[0014] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0015] To improve the control efficiency of offline transactions, this specification provides an audio processing method. The subject of this audio processing method is a processing device or a mouthpiece device with a processing device. The following is a detailed description. It should be noted that the order of description of the following embodiments is not intended to limit the preferred order of the embodiments.

[0016] Please see Figure 1 , Figure 1 This is a schematic diagram illustrating a scenario of the audio processing method provided in the embodiments of this specification. For example... Figure 1 As shown, the mouth-worn device consists of a shell 10, a breathing valve 11, an external speaker 12, a sound-absorbing cavity 14, and an active noise cancellation device 15. Specifically, the shell 10 can be a rigid carbon fiber shell, with a recessed area inside forming an accommodating space. The breathing valve 11 is installed on the shell 10 to connect the accommodating space with the external environment. A second rubber sealing ring is fitted onto the area of ​​the shell 10 that contacts the user's face to ensure airtightness when the shell contacts the face. The sound-absorbing cavity 14 is located within this accommodating space and directly covers the user's mouth area. The active noise cancellation device 15 is embedded in the inner wall of the sound-absorbing cavity 14 and generates a reverse sound wave signal for active noise reduction. A second microphone 17 is installed on the shell 10 to collect audio signals generated by the user within the accommodating space; specifically, the second microphone 17 is an omnidirectional microphone. An external speaker 12 is also installed on the shell 10 to play the audio signals outwards. The mouthpiece is secured to the user's ear by an elastic cord 20. The elastic cord 20 is also equipped with a tension adjuster 21 to adjust the tightness of the fit, and is combined with a second rubber sealing ring 19 to ensure the airtightness of the housing space and prevent sound leakage.

[0017] Please refer to the following: Figure 2 , Figure 2This is a schematic diagram illustrating a scenario of the audio processing method provided in the embodiments of this specification. For example... Figure 2 The side view of the anechoic chamber 14 shown is D-shaped and consists of an inner cavity and an outer cavity, with sound-absorbing foam 22 filling the space between the inner and outer cavities. The active anechoic device 15 consists of a first microphone 16 and an external speaker 13. Specifically, the first microphone 16 is a unidirectional microphone. The first microphone 16 and the external speaker 13 are installed in preset positions within the anechoic chamber 14. Figure 2 The installation positions of the first microphone 16 and the reverse speaker 13 are only one preferred embodiment. Based on the above description, it is also reasonable to directly and meaninglessly deduce that the first microphone 16 and the reverse speaker 13 can be installed in other positions of the anechoic cavity 14, which will not be elaborated here. The first microphone 16 is used to collect the audio signal of the user, and the reverse speaker 13 is used to generate a reverse sound wave signal. A first rubber sealing ring 18 is provided on the left edge of the anechoic cavity 14, which contacts the user's mouth area, to ensure the airtightness of the anechoic cavity 14 when in contact with the mouth.

[0018] Please refer to the following: Figure 3 , Figure 3 This is a schematic diagram illustrating a scenario of the audio processing method provided in the embodiments of this specification. For example... Figure 3 The diagram shows a front view of the mouth-worn device. The device has an overall approximately spherical shape. A second microphone 17 is located at the top of the device's housing for collecting audio signals from within the housing. An active noise cancellation device 15, integrating a first microphone 16 and a reverse speaker 13, is located at the center of the front of the device. A processing unit 23 is located at the bottom of the device. This processing unit 23 integrates a microcontroller unit, a voice calculation model unit, and a digital signal processor, and also includes a built-in Bluetooth wireless communication module for connecting to the user's mobile device.

[0019] Traditional passive noise cancellation devices rely on multiple layers of sound-absorbing materials to attenuate sound wave energy, resulting in a significant increase in device size and weight, affecting wearing comfort and aesthetics. During sound wave propagation, low-frequency components, due to their longer wavelengths, penetrate the sound-absorbing materials and still couple with the air medium through structural vibrations to form secondary radiation, causing residual effective sound pressure levels and limiting acoustic privacy protection performance. For example, in an open-plan office setting, when users wear traditional noise-canceling masks for conference calls, the internal acoustic cavity of the device needs to accommodate four layers of composite sound-absorbing layers to achieve noise reduction in frequencies above 3000Hz. However, the increased thickness of the device (25mm) leads to pressure on the face. When a user emits speech containing a 200Hz fundamental frequency, the sound waves penetrate the sound-absorbing layers, causing resonance of 0.5mm amplitude in the mask shell. This resonance propagates through the air to a distance of 1 meter, creating a sound pressure level of 45dB, resulting in leakage of speech content.

[0020] If the above issues are not addressed, the leakage of users' voice information will increase the risk of users' privacy data being compromised.

[0021] Based on the aforementioned problems, this application first considers the inherent drawback of traditional passive noise reduction devices, which rely on the stacking of physical materials, resulting in a bulky structure. While multi-layered sound-absorbing materials can attenuate high-frequency sound waves, they cannot effectively suppress structural vibrations and secondary radiation caused by low-frequency sound waves. Therefore, this application reduces the number of material layers while maintaining noise reduction effectiveness. Further analysis reveals that actively generating reverse sound waves can cancel out the energy of the original sound waves, eliminating the need for material thickness to attenuate sound waves. Low-frequency sound waves are more easily eliminated through phase interference due to their wavelength characteristics. Based on this, this application chooses to embed an active noise cancellation device within the anechoic cavity, directly interfering with the sound wave propagation path by predicting the user's voice signal in real time and generating reverse sound waves. The following will combine... Figures 4-10 This document provides a detailed description of an audio processing method provided in the embodiments of this specification.

[0022] Please see Figure 4 , Figure 4 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification. Figure 4 As shown, the method in the embodiments of this specification may include the following steps S102-S106.

[0023] S102, when the active noise cancellation function of the mouth-worn device is enabled, acquire the first audio signal of the wearer within a preset duration.

[0024] In the embodiments described in this specification, the mouth-worn device can be paired and connected with the user's mobile device, and the user can control the mouth-worn device to activate the active noise cancellation function through the mobile device.

[0025] It should be noted that the first audio signal refers to the audio data emitted by the wearer within a very short historical period, such as the audio data from microseconds or nanoseconds ago, obtained through a one-way microphone after the active noise cancellation function is activated.

[0026] Specifically, when the active noise cancellation function of the mouth-worn device is enabled, the unidirectional microphone integrated in the active noise cancellation device is activated. The unidirectional microphone is oriented towards the internal space of the noise cancellation cavity to collect audio signals generated by the user's mouth and directly propagating within the cavity in real time. The processing device writes the audio signals into a dedicated data storage area, which is configured to dynamically save audio data within a preset historical time period. When the predictive noise cancellation algorithm needs to be activated, the audio data within the preset historical time period is read from the data storage area and extracted, and defined as the first audio signal.

[0027] Please refer to the following: Figure 5 , Figure 5 This is a schematic diagram illustrating a scenario of an audio processing method provided in an embodiment of this specification. For example... Figure 5 As shown, the user's mobile device contains an application for controlling the mouthpiece device, specifically a "mouthpiece device controller". In this application, the "active noise cancellation function" is enabled, the "active sound enhancement function" is disabled, and there is a "preset configuration information adjustment" button for users to personalize their settings. The bottom of the interface clearly displays the real-time status of the device. The left side shows a "connected" prompt, indicating that the mouthpiece device has been paired and connected to the mobile device, while the right side shows that the mouthpiece device currently has 90% battery remaining.

[0028] S104, perform audio feature analysis on the first audio signal to obtain the predicted audio signal, and call the active noise cancellation device to generate the reverse sound wave signal corresponding to the predicted audio signal.

[0029] Specifically, when the active noise cancellation function of the mouth-worn device is enabled, the processing device first performs audio feature analysis on the first audio signal. Audio feature analysis is performed by calling a preset speech prediction model to extract the audio acoustic features and phoneme sequence of the first audio signal, and generating a predicted phoneme sequence based on these features. The predicted phoneme sequence is then converted into continuous waveform data to obtain the predicted audio signal, that is, the predicted speech waveform of the user at a future moment.

[0030] The processing device invokes an active silencing device, which consists of a unidirectional microphone and a reverse speaker, and is embedded in the silencing cavity to generate a reverse acoustic wave signal corresponding to the predicted audio signal. The generation process includes determining the degradation equation of the predicted audio signal in the path through the silencing cavity and the outer shell, performing audio degradation processing on the predicted audio signal based on the degradation equation to obtain a degraded audio signal, and then performing signal inversion processing on the degraded audio signal to generate a reverse acoustic wave signal with equal amplitude and completely opposite phase.

[0031] S106, invoke the active silencing device to output a reverse acoustic wave signal.

[0032] Specifically, after obtaining the reverse acoustic wave signal based on the above step S104, the processing device identifies the reverse loudspeaker in the active silencing device as the execution unit and outputs the reverse acoustic wave signal through the reverse loudspeaker.

[0033] Please refer to the following: Figure 6 , Figure 6 This is a schematic diagram illustrating a scenario of an audio processing method provided in an embodiment of this specification. For example... Figure 6 As shown, when the user activates the active noise cancellation function of the mouthpiece via a mobile device, the unidirectional microphone of the mouthpiece sends the first audio signal emitted by the user to the speech prediction model built into the processing device. The speech prediction model, after calling a preset noise cancellation algorithm, obtains the predicted audio signal corresponding to the first audio signal. Further, by performing reverse processing on the predicted audio signal, the processing device obtains the corresponding reverse sound wave signal. The processing device designates the reverse speaker as the execution unit and outputs the reverse sound wave signal through the reverse speaker. Simultaneously, the user can control the mouthpiece to deactivate the active noise cancellation function via their mobile device.

[0034] In the embodiments described in this specification, by predicting the user's speech in real time and generating a reverse sound wave, the propagation path of the sound wave is directly intervened, achieving efficient noise cancellation without relying on multi-layer material stacking. Especially for low-frequency sound waves, the principle of phase interference is used for cancellation, overcoming the problem of poor noise cancellation in the low-frequency band of traditional passive noise cancellation devices. At the same time, by reducing the use of physical sound-absorbing materials, the thickness and weight of the device are significantly reduced, improving wearing comfort and aesthetics. Furthermore, by combining active and passive noise cancellation, effective noise reduction is achieved across the entire frequency band, significantly improving voice privacy protection.

[0035] In some embodiments described above, when the active noise cancellation device generates the reverse sound wave signal corresponding to the predicted audio signal, the predicted signal may deviate from the actual sound field due to environmental noise, device delay, or signal attenuation during propagation. Consequently, the reverse sound wave cannot accurately cancel the target sound wave, affecting the noise cancellation effect. Therefore, please refer to [link to relevant documentation]. Figure 7 , Figure 7 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification. Figure 7 As shown, the method in the embodiments of this specification may include the following steps S202-S206.

[0036] S202, invoke the active noise cancellation device to determine the signal degradation equation of the predicted audio signal.

[0037] In this embodiment of the application, the predicted audio signal is the audio signal that the user will generate after a preset future time.

[0038] Specifically, the predicted audio signal is acquired and its distortion and attenuation characteristics after penetrating the anechoic cavity and the device casing are analyzed. This includes, but is not limited to, reductions in audio signal amplitude and frequency distortion. A signal degradation equation is derived by comprehensively modeling based on this distortion and attenuation data. This equation describes how the predicted audio signal degrades into residual noise that eventually leaks into the external environment.

[0039] S204, determine the inverse acoustic signal corresponding to the predicted audio signal based on the signal degradation equation.

[0040] Specifically, after obtaining the signal degradation equation in step S202 above, the processing device will perform inverse mathematical calculations based on the signal degradation equation to derive the reverse acoustic wave signal. When the reverse acoustic wave signal and the predicted audio signal leaked after passing through the path described by the degradation equation are superimposed in space, the energy of the two can be effectively canceled through the complete phase inversion of the waveform and the precise matching of the amplitude, thereby achieving the purpose of neutralizing and eliminating the leakage noise.

[0041] S206, calls the reverse speaker to output a reverse sound wave signal.

[0042] Specifically, after determining the reverse acoustic wave signal, the processing device generates a corresponding drive command and sends it to the reverse loudspeaker in the active anechoic chamber. The reverse loudspeaker, acting as the execution unit, converts the received digital signal into an analog electrical signal, which in turn drives its diaphragm to generate a reverse acoustic wave signal that is completely opposite in phase and corresponding in amplitude to the predicted leaking acoustic wave. This reverse acoustic wave is then released into the internal space of the anechoic chamber to actively intervene in and cancel out the audio signal that is about to leak outwards.

[0043] In the embodiments described in this specification, an inverse sound wave signal with the opposite phase to the predicted audio signal can be accurately generated and precisely output through an anti-phase loudspeaker. This active noise cancellation method can effectively cancel out sounds across a wide frequency range, significantly improving the noise cancellation effect. Furthermore, due to the use of prediction and signal processing techniques, it can quickly respond to changes in sound, achieving real-time dynamic noise cancellation.

[0044] In some embodiments described above, when generating the reverse acoustic wave signal based on the signal degradation equation, a time difference exists between the predicted audio signal and the actual emitted sound signal. This causes the reverse acoustic wave to not completely cancel out the original sound wave, affecting the noise reduction effect. Please refer to... Figure 8 , Figure 8 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification. Figure 8 As shown, the method in the embodiments of this specification may include the following steps S302-S304.

[0045] S302, based on the degradation equation, performs audio degradation processing on the predicted audio signal to obtain the degraded audio signal.

[0046] In the embodiments described in this specification, the predicted audio signal is the audio signal predicted to be generated by the user after a preset future time. Specifically, the speech prediction model can be a model generated based on a deep neural network structure.

[0047] Specifically, the real-time acquired or pre-generated current predicted audio signal is input into a preset speech prediction model. The input signal undergoes high-precision audio feature analysis to extract key acoustic features such as Mel spectrum, fundamental frequency, and phoneme duration. Based on these features and the learned speech production rules, the model then performs temporal extrapolation and regression prediction on the pronunciation trend of the wearer in the next short period of time (i.e., a preset duration, such as 50-200 milliseconds), thereby generating predicted acoustic parameters representing the sound characteristics at future moments. Finally, the model or the backend vocoder combines these predicted parameters into a complete time-domain waveform, which is the predicted audio signal corresponding to the preset duration.

[0048] S304 performs signal inversion processing on the degraded audio signal to obtain an inverted sound wave signal.

[0049] Specifically, the degradation equation calculates the sound wave reflection path based on the cavity's geometric parameters and superimposes the ambient noise spectrum to generate a degradation signal. The phase reversal algorithm adjusts the phase of the degradation signal by 180 degrees, generating an inverse sound wave signal with equal amplitude and opposite polarity.

[0050] In the embodiments described in this specification, accurate prediction and active noise cancellation of the wearer's future speech are achieved. By employing a speech prediction model, reverse sound waves can be generated in advance, eliminating the delay problem of traditional active noise cancellation systems. Simultaneously, through the introduction of a signal degradation equation, the reverse sound waves more closely match the actual propagation environment, improving the noise cancellation effect. This method not only effectively eliminates high-frequency speech components but also suppresses the propagation of low-frequency sounds, significantly improving the overall noise cancellation effect.

[0051] In some embodiments described above, a method is proposed to analyze the audio features of a predicted audio signal using a speech prediction model to generate a predicted phoneme sequence. However, in this process, the lack of joint analysis of the audio acoustic features and the phoneme sequence leads to insufficient accuracy in the predicted audio signal, thus affecting the generation of the reverse acoustic wave signal. Therefore, please refer to [link to relevant documentation]. Figure 9 , Figure 9 This is a flowchart illustrating an audio processing method provided in an embodiment of this specification. Figure 9 As shown, the method in the embodiments of this specification may include the following steps S402-S406.

[0052] S402, based on the speech prediction model, obtain the audio acoustic features and phoneme sequence of the first audio signal.

[0053] In the embodiments of this specification, the speech prediction model can employ a deep neural network structure, including convolutional neural network layers and recurrent neural network layers. The convolutional neural network layers are used to extract local features of the audio signal, while the recurrent neural network layers are used to capture the temporal dependencies of the audio signal. By training on a large amount of speech data, the speech prediction model can learn the mapping relationship between the acoustic features and linguistic features of speech, thereby extracting the spectral, fundamental frequency, or formant features of the audio signal. Phoneme sequence generation can be combined with a hidden Markov model or an end-to-end phoneme recognition model to map the audio acoustic features into a phoneme probability distribution.

[0054] S404 generates a predicted phoneme sequence based on audio acoustic features and phoneme sequences.

[0055] Specifically, the input first audio signal is segmented into frames using a speech prediction model, and the Mel-frequency cepstral coefficients of each frame are extracted as audio acoustic features. Based on the time series of the acoustic features, the speech prediction model outputs a corresponding phoneme label sequence, for example, by using a connection-based temporal classification algorithm to achieve a frame-level to phoneme-level mapping. The generated phoneme sequence and the original acoustic features are input together into the phoneme prediction module, which uses a bidirectional long short-term memory network to capture contextual dependencies and outputs a predicted phoneme sequence after a preset duration.

[0056] S406 converts the predicted phoneme sequence into an audio signal to obtain the predicted audio signal.

[0057] In the embodiments described in this specification, the predicted audio signal is the audio signal generated by the user after a preset future time.

[0058] Specifically, the processing device takes the predicted phoneme sequence as input, converts each phoneme and its context information into a corresponding acoustic feature sequence, such as a Mel spectrogram, through a speech prediction model, decodes the acoustic feature sequence through a built-in vocoder, and reconstructs it into continuous, time-domain waveform data through waveform generation technology. Finally, after necessary signal processing such as digital-to-analog conversion, the output is a predicted audio signal that can be directly played by a speaker.

[0059] In the embodiments described in this specification, accurate prediction of a user's speech content over a future period can be achieved, providing a precise input signal for subsequent active noise cancellation. This improves the effectiveness and real-time performance of active noise cancellation, enabling the wearable device to better protect the user's voice privacy. Furthermore, by employing a deep learning model for speech prediction, the system can adapt to the different speech characteristics and speaking habits of various users, improving its versatility and robustness.

[0060] In some embodiments described above, the active noise cancellation function eliminates the audio signal in the user's mouth area by using a reverse sound wave signal, but it does not address how to effectively collect and output the user's voice, resulting in a lack of corresponding audio processing mechanisms when the user needs to actively speak. Therefore, in a feasible embodiment, the method of this specification may further include the following steps: When the active speaker function of the mouth-worn device is enabled, the second audio signal of the wearer is acquired.

[0061] In the embodiments described in this specification, the second audio signal is acquired through a first microphone and / or a second microphone on the housing of the mouth-worn device. An external speaker is located on the surface of the housing, and its acoustic directivity is designed to face the environment outside the user's mouth.

[0062] The second audio signal is output through the external speaker.

[0063] Specifically, when a user controls the mouthpiece to activate the active speaker function via a mobile device, the second microphone collects the voice vibration signal of the user's mouth area in real time within the anechoic chamber, generates a second audio signal, and outputs the second audio signal to the external environment through an external speaker.

[0064] In the embodiments described in this specification, the user's privacy is protected while ensuring that the user's voice is clearly transmitted to the external environment. When the user needs to communicate with others, there is no need to remove the mask; simply activating the active speaker function allows the other party to hear the user's voice clearly. This method protects the user's health and safety while improving the convenience and efficiency of communication. Furthermore, due to the use of active speaker technology, compared to traditional masks where users need to raise their voices to be heard, the solution in this application allows users to speak at a normal volume, reducing the user's vocal burden and improving comfort.

[0065] In some embodiments described above, when the active speaker function is enabled and the second audio signal is directly output, the user's personalized audio output needs are not considered. This results in the audio signal not being optimized, which may affect the output sound quality and clarity, and fail to meet the usage requirements in different scenarios. Therefore, in a feasible embodiment, when performing the above steps to output the second audio signal via an external speaker, the following is also specifically performed: Obtain the user's preset configuration information for the active speaker function.

[0066] The second audio signal is optimized based on the preset configuration information, and the optimized second audio signal is output through the external speaker.

[0067] In the embodiments described in this specification, the preset configuration information consists of volume parameters, frequency band gain parameters, or noise reduction level parameters pre-set by the user. The preset configuration information is stored in the user's mobile device, which is connected to the processing device in the mouth-worn device. Specifically, when the user activates the active speaker function of the mouthpiece via a mobile device, the processing device continuously collects the user's voice signal through a second microphone, which is an omnidirectional microphone, to form a second audio signal. The processing device then sends the second audio signal to the mobile device, which reads user-preset configuration parameters stored in the user's device. These parameters include a pre-set volume curve, frequency band equalization parameters, and background noise suppression level. The mobile device performs at least frequency band gain and noise reduction processing on the second audio signal based on the pre-set configuration information to obtain an optimized second audio signal. The mobile device then sends the second audio signal back to the processing device, which then uses an external speaker to output the optimized second audio signal.

[0068] For example, the dynamic range of the second audio signal is compressed from 80dB to 60dB. The frequency band gain parameter is increased by 3dB in the 200Hz-500Hz band and attenuated by 2dB in the 2kHz-4kHz band.

[0069] Please refer to the following: Figure 10 , Figure 10 This is a schematic diagram illustrating a scenario of an audio processing method provided in an embodiment of this specification. For example... Figure 10 As shown, when the user activates the active speaker function of the mouthpiece via a mobile device, the processing unit of the mouthpiece controls the unidirectional and / or omnidirectional microphones to send the second audio signal emitted by the user to the mobile device. The mobile device, after obtaining the user's preset configuration information for the active speaker function, adjusts and reduces the noise of the second audio signal before sending the optimized second audio signal to the processing unit of the mouthpiece. The processing unit uses the external speaker as an execution unit and outputs the optimized second audio signal to the outside through the external speaker. Simultaneously, the user can control the mouthpiece to deactivate the active speaker function via the mobile device.

[0070] In the embodiments described in this specification, personalized customization of the output audio signal is achieved. Users can flexibly adjust the audio output effect according to their personal needs and usage scenarios, improving the user experience. Simultaneously, through optimized processing of the audio signal, the quality and clarity of the output audio are improved, enabling the wearer's voice to be better transmitted to the external environment. Furthermore, this solution can adjust audio characteristics according to different scenario requirements; for example, enhancing mid-to-high frequencies in noisy environments to improve voice clarity, and increasing low frequencies in quiet environments to improve sound fullness, thus adapting to various usage scenarios.

[0071] When all active noise cancellation functions are enabled, the user's audio signal will be output through the external speaker. However, some audio signals will penetrate the noise cancellation cavity and shell and diffuse outwards, causing sound interference and affecting the listening experience. Therefore, in a feasible embodiment, the method of this specification may further include the following steps: When both the active speaker function and the active noise cancellation function of the mouth-worn device are enabled, the reverse sound wave signal of the predicted audio signal is output based on the reverse speaker, and the third audio signal of the wearer is output based on the external speaker.

[0072] Specifically, the processing device acquires the third audio signal generated by the user's mouth in real time through a unidirectional microphone set in the anechoic cavity, performs feature analysis on the signal to predict its waveform at future moments, and generates a corresponding predicted audio signal. Subsequently, the processing device calls the active anechoic device to determine the mathematical model of amplitude attenuation and frequency distortion of the predicted audio signal during its penetration of the anechoic cavity and shell, namely the signal degradation equation. Based on this equation, the device performs audio degradation processing on the predicted audio signal to obtain a degraded audio signal simulating a leakage sound field. Then, the device performs signal inversion processing on the degraded audio signal to generate an inverse sound wave signal with the same amplitude but completely opposite phase.

[0073] The processing device controls the reverse speaker embedded in the anechoic cavity to output the reverse sound wave signal, so that it interferes and cancels out the actual leaked sound wave inside the cavity, thereby suppressing the sound from spreading outward; and outputs the collected third audio signal to the external environment through the external speaker set on the outer shell, thereby realizing the parallel processing of sound wave cancellation and voice playback, effectively solving the problem of mutual interference caused by audio signal leakage, and significantly improving the purity and listening quality of the output voice.

[0074] In one feasible implementation, based on the existing hardware architecture and audio processing capabilities of the mouth-worn device, functions such as real-time translation, cross-language teaching, and fun voice changing can be deeply integrated by expanding the software algorithms of the processing device and integrating with cloud services.

[0075] Specifically, while maintaining active noise cancellation and speaker output, the processing unit loads a multi-task speech model via its built-in NPU and establishes a low-latency communication link with mobile devices or cloud servers using a Bluetooth module. For real-time translation, when the user speaks, the raw audio captured by the unidirectional microphone is converted into text in real-time by the Neural Processing Unit (NPU), then converted into the target language by a cloud-based translation engine, and finally synthesized into natural speech and output through external speakers. The cross-language teaching function adds a pronunciation comparison module, which analyzes the differences between the user's phonemes and the standard library using the NPU, and provides focused auditory feedback using the acoustic isolation characteristics of the noise-canceling cavity, while simultaneously playing corrective demonstrations through external speakers. It also supports real-time modification of the fundamental frequency and formant characteristics of the audio signal via a Digital Signal Processor (DSP), supports switching between multiple preset sound effect modes, and directly outputs the voice-changing results through external speakers. All functions are managed uniformly through a mobile application, allowing users to dynamically switch modes. The processing unit can intelligently allocate computing resources according to the current scenario, ensuring real-time performance and stability when multiple functions are running in parallel.

[0076] based on Figures 1-3 The following is a detailed description of the oral wearing device provided in the embodiments of this specification, illustrated in the scenario diagram. It should be noted that the oral wearing device is used to perform the functions described above in this specification. Figures 1-10 The embodiments and methods described herein are for illustrative purposes only, showing the parts relevant to the embodiments of this specification. For specific technical details not disclosed, please refer to this specification. Figures 1-10 In the embodiment shown, the mouth-worn device includes a processing device 23, a housing 10, a breathing valve 11, an external speaker 12, a sound-absorbing cavity 14, and an active sound-absorbing device 15. A recessed area in the housing 10 forms an accommodating space. The breathing valve 1111 is disposed on the housing 10 to connect the accommodating space with the ambient space outside the housing 10. The external speaker 12 is disposed on the housing 10 to output audio signals within the accommodating space. The sound-absorbing cavity 14 is disposed within the accommodating space to cover the mouth area of ​​the wearer. The active sound-absorbing device 15 is embedded within the cavity of the sound-absorbing cavity 14. Processing device 23 is used to acquire the first audio signal of the wearer within a preset duration when the active noise cancellation function of the mouth-worn device is enabled. Audio feature analysis is performed on the first audio signal to obtain the predicted audio signal; The active noise cancellation device 15 is invoked to generate an inverse acoustic wave signal corresponding to the predicted audio signal. The inverse acoustic wave signal is an audio signal with the opposite phase to the predicted audio signal. The active silencing device 15 is activated to output a reverse acoustic wave signal.

[0077] In some embodiments, the processing device 23 is further configured to perform the following steps: The active noise cancellation device 15 is invoked to determine the signal degradation equation of the predicted audio signal; The reverse acoustic wave signal corresponding to the predicted audio signal is determined based on the signal degradation equation. The inverting speaker outputs a reverse sound wave signal.

[0078] In some embodiments, the processing device 23 is further configured to perform the following steps: The predicted audio signal is subjected to audio degradation processing based on the degradation equation to obtain the degraded audio signal; The degraded audio signal is inverted to obtain an inverted sound wave signal. The inverting speaker outputs a reverse sound wave signal.

[0079] In some embodiments, the processing device 23 is further configured to perform the following steps: The predicted audio signal is subjected to audio degradation processing based on the degradation equation to obtain the degraded audio signal; The degraded audio signal is inverted to obtain an inverted sound wave signal.

[0080] In some embodiments, the processing device 23 is further configured to perform the following steps: The audio acoustic features and phoneme sequence of the first audio signal are obtained based on a speech prediction model. Based on audio acoustic features and phoneme sequences, a predicted phoneme sequence is generated; The predicted phoneme sequence is converted into an audio signal to obtain the predicted audio signal.

[0081] In some embodiments, the processing device 23 is further configured to perform the following steps: Based on audio acoustic features and phoneme sequences, a predicted phoneme sequence is generated; The predicted phoneme sequence is converted into an audio signal to obtain the predicted audio signal.

[0082] In some embodiments, the processing device 23 is further configured to perform the following steps: When the active speaker function of the mouth-worn device is enabled, acquire the second audio signal from the user. The second audio signal is output based on the external speaker 12. In some embodiments, the processing device 23 is further configured to perform the following steps: Obtain the user's preset configuration information for the active speaker function; The second audio signal is optimized based on the preset configuration information, and the optimized second audio signal is output through the external speaker 12.

[0083] In some embodiments, the processing device 23 is further configured to perform the following steps: When both the active sound-speaking function and the active sound-canceling function of the mouth-worn device are enabled, the reverse sound wave signal of the predicted audio signal is output based on the reverse speaker 13, and the third audio signal of the wearer is output based on the external speaker 12.

[0084] In some embodiments, the active noise cancellation device 15 is provided with a first microphone 16 and a reverse speaker, wherein the first microphone 16 is a unidirectional microphone.

[0085] In some embodiments, the housing 10 is provided with a second microphone 17, which is an omnidirectional microphone.

[0086] In some embodiments, the anechoic cavity 14 consists of an inner cavity and an outer cavity, and the space between the inner cavity and the outer cavity is filled with sound-absorbing foam 22.

[0087] In some embodiments, a first rubber sealing ring 18 is provided at the edge of the cavity in the anechoic chamber 14 that contacts the mouth area of ​​the wearer, and a second rubber sealing ring 19 is provided at the shell area in the outer shell 10 that contacts the face area of ​​the wearer.

[0088] In some embodiments, the mouth-worn device includes a battery, which is electrically connected to a processing device, an external speaker, an active noise cancellation device, a first microphone 16, and a second microphone.

[0089] In the embodiments described in this specification, by predicting the user's speech in real time and generating a reverse sound wave, the propagation path of the sound wave is directly intervened, achieving efficient noise cancellation without relying on multi-layer material stacking. Especially for low-frequency sound waves, the principle of phase interference is used for cancellation, overcoming the problem of poor noise cancellation in the low-frequency band of traditional passive noise cancellation devices. At the same time, by reducing the use of physical sound-absorbing materials, the thickness and weight of the device are significantly reduced, improving wearing comfort and aesthetics. Furthermore, by combining active and passive noise cancellation, effective noise reduction is achieved across the entire frequency band, significantly improving voice privacy protection.

[0090] Furthermore, the mouth-worn device provided in the above embodiments and the audio processing method embodiment belong to the same concept, and the implementation process can be found in the method embodiment, which will not be repeated here.

[0091] The embodiment numbers in this specification are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. In some cases, the actions or steps described in the claims can be performed in a different order than that shown in the embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0092] Please see Figure 11 This document provides a schematic diagram of the structure of an electronic device as an embodiment of the present specification. Figure 11 As shown, the electronic device 500 includes a processor 501 and a memory 502. The processor 501 and the memory 502 are electrically connected.

[0093] Processor 501 is the control center of electronic device 500 and may include one or more processing cores. Processor 501 connects to various parts of the electronic device using various interfaces and lines. By running or calling computer programs stored in memory 502, and by calling data stored in memory 502, it executes various functions and processes data of the electronic device, thereby providing overall control over the electronic device. Optionally, processor 501 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). Processor 501 may integrate one or more of the following: processor, Graphics Processing Unit (GPU), and modem. The processor mainly handles the operating system, user page, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem is used for wireless communication. It is understood that the modem may also not be integrated into processor 501 and may be implemented separately through a communication chip.

[0094] The memory 502 can be used to store software programs and modules. The processor 501 executes various functional applications and controls offline tasks by running the computer programs and modules stored in the memory 502. The memory 502 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, computer programs required for at least one function, etc.; the data storage area may store data created according to the use of the electronic device, etc.

[0095] Furthermore, memory 502 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, memory 502 may also include a memory controller to provide processor 501 with access to memory 502.

[0096] In the embodiments described in this specification, the processor 501 in the electronic device 500 loads the instructions corresponding to the processes of one or more computer programs into the memory 502 according to the following steps, and the processor 501 runs the computer programs stored in the memory 502 to realize various functions, as follows: When the active noise cancellation function of the mouth-worn device is enabled, the first audio signal of the wearer within a preset duration is acquired. The first audio signal is subjected to audio feature analysis to obtain the predicted audio signal. The active noise cancellation device is called to generate the reverse sound wave signal corresponding to the predicted audio signal. The reverse sound wave signal is an audio signal with the opposite phase to the predicted audio signal. The active silencing device is activated to output a reverse acoustic wave signal.

[0097] Optionally, the processor 501 executes the following steps when invoking the active noise cancellation device to generate the inverse acoustic wave signal corresponding to the predicted audio signal: The active noise cancellation device is invoked to determine the signal degradation equation of the predicted audio signal; The reverse acoustic wave signal corresponding to the predicted audio signal is determined based on the signal degradation equation. The inverting speaker outputs a reverse sound wave signal.

[0098] Optionally, the processor 501 executes the following steps when determining the inverse acoustic signal corresponding to the predicted audio signal based on the signal degradation equation: The predicted audio signal is subjected to audio degradation processing based on the degradation equation to obtain the degraded audio signal; The degraded audio signal is inverted to obtain an inverted sound wave signal.

[0099] Optionally, the processor 501 performs audio feature analysis on the first audio signal to obtain a predicted audio signal, specifically executing: The audio acoustic features and phoneme sequence of the first audio signal are obtained based on the speech prediction model; Based on audio acoustic features and phoneme sequences, a predicted phoneme sequence is generated; The predicted phoneme sequence is converted into an audio signal to obtain the predicted audio signal.

[0100] Optionally, processor 501 is also used for specific execution: When the active speaker function of the mouth-worn device is enabled, acquire the second audio signal from the user. The second audio signal is output through the external speaker.

[0101] Optionally, the processor 501, when executing the function of outputting a second audio signal based on the external speaker, specifically performs the following: Obtain the user's preset configuration information for the active speaker function; The second audio signal is optimized based on the preset configuration information, and the optimized second audio signal is output through the external speaker.

[0102] Optionally, the processor 501 is also used to specifically perform: when both the active sound-speaking function and the active sound-canceling function of the mouth-worn device are enabled, outputting the reverse sound wave signal of the predicted audio signal based on the reverse speaker, and outputting the third audio signal of the wearer based on the external speaker.

[0103] In the embodiments described in this specification, user speech is predicted in real time and an inverse sound wave is generated, directly interfering with the sound wave propagation path to achieve efficient noise reduction without relying on multi-layer material stacking. Especially for low-frequency sound waves, the principle of phase interference is used for elimination, overcoming the problem of poor noise reduction in the low-frequency band of traditional passive noise reduction devices. At the same time, by reducing the use of physical sound-absorbing materials, the thickness and weight of the device are significantly reduced, improving wearing comfort and aesthetics. Furthermore, by combining active and passive noise reduction, effective noise reduction is achieved across the entire frequency band, significantly improving voice privacy protection.

[0104] In addition, the processing device provided in the embodiments of this specification may specifically be a chip, component or module. The chip may include a connected processor and a memory. The memory is used to store instructions. When the processor calls and executes the instructions, the chip can execute an audio processing method provided in the above embodiments.

[0105] This specification also provides a computer-readable storage medium storing computer program code. When the computer program code is run on a computer, the computer executes the aforementioned method steps to implement the audio processing method provided in the above embodiments.

[0106] This specification also provides a computer program product that stores at least one instruction, which is loaded and executed by a processor as described above. Figures 4-10 The audio processing method of the illustrated embodiment can be found in the following document for detailed execution process. Figures 4-10 The specific details of the illustrated embodiments will not be elaborated here.

[0107] The apparatus, computer-readable storage medium, computer program product or chip provided in the embodiments of this specification are all used to execute the corresponding methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0108] Through the above description of the embodiments, those skilled in the art will understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0109] In the embodiments provided in this specification, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the related couplings or direct couplings or communication connections shown or discussed may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0110] The above description is merely a specific embodiment of this specification, but the scope of protection of this specification is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this specification should be included within the scope of protection of this specification. Therefore, the scope of protection of this specification should be determined by the scope of the claims.

Claims

1. An audio processing method, characterized in that, An application is made to a processing device used in a mouth-worn device. The mouth-worn device includes a housing, a breathing valve, an external speaker, a sound-absorbing cavity, and an active sound-absorbing device. A recessed area in the housing forms an accommodating space. The breathing valve is disposed on the housing to connect the accommodating space with the ambient space outside the housing. The external speaker is disposed on the housing for collecting audio signals within the accommodating space. The sound-absorbing cavity is disposed within the accommodating space to cover the mouth area of ​​the wearer. The active sound-absorbing device is embedded within the cavity of the sound-absorbing cavity. The method includes: When the active noise cancellation function of the mouth-wearing device is enabled, the first audio signal of the wearer within a preset duration is acquired. The first audio signal is subjected to audio feature analysis to obtain a predicted audio signal. The active noise cancellation device is then invoked to generate a reverse acoustic wave signal corresponding to the predicted audio signal. The reverse acoustic wave signal is an audio signal with the opposite phase to the predicted audio signal. The active noise cancellation device is invoked to output the reverse acoustic wave signal.

2. The method according to claim 1, wherein the active noise cancellation device comprises a first microphone and a reverse speaker; The step of invoking the active noise cancellation device to generate the inverse acoustic wave signal corresponding to the predicted audio signal includes: The active noise cancellation device is invoked to determine the signal degradation equation of the predicted audio signal; The inverse acoustic signal corresponding to the predicted audio signal is determined based on the signal degradation equation. The reverse speaker is invoked to output the reverse sound wave signal.

3. The method according to claim 2, wherein determining the inverse acoustic signal corresponding to the predicted audio signal based on the signal degradation equation comprises: The predicted audio signal is subjected to audio degradation processing based on the degradation equation to obtain a degraded audio signal; The degraded audio signal is inverted to obtain an inverted sound wave signal.

4. The method according to claim 1, wherein performing audio feature analysis on the first audio signal to obtain a predicted audio signal comprises: The audio acoustic features and phoneme sequence of the first audio signal are obtained based on the speech prediction model; Based on the audio acoustic features and the phoneme sequence, a predicted phoneme sequence is generated; The predicted phoneme sequence is converted into an audio signal to obtain the predicted audio signal.

5. The method according to claim 1, wherein the housing includes a second microphone, and the method further includes: When the active speaker function of the mouth-worn device is enabled, the second audio signal of the wearer is acquired; The second audio signal is output based on the external speaker.

6. The method according to claim 5, wherein outputting the second audio signal based on the external speaker comprises: Obtain the user's preset configuration information for the active speaker function; The second audio signal is optimized based on the preset configuration information, and the optimized second audio signal is output through the external speaker.

7. The method according to claim 5, further comprising: When both the active sound-speaking function and the active sound-canceling function of the mouth-worn device are enabled, the reverse sound wave signal of the predicted audio signal is output based on the reverse speaker, and the third audio signal of the wearer is output based on the external speaker.

8. A mouth-worn device, the mouth-worn device comprising a processing device, a housing, a breathing valve, an external speaker, a sound-absorbing cavity, and an active sound-absorbing device, wherein a recessed area in the housing forms an accommodating space, the breathing valve is disposed on the housing to connect the accommodating space with the ambient space outside the housing, the external speaker is disposed on the housing to output an audio signal within the accommodating space, the sound-absorbing cavity is disposed within the accommodating space to cover the mouth area of ​​the wearer, and the active sound-absorbing device is embedded within the cavity of the sound-absorbing cavity; The processing device is used to acquire the first audio signal of the wearer within a preset duration when the active noise cancellation function of the mouth-wearing device is enabled. The first audio signal is subjected to audio feature analysis to obtain the predicted audio signal; The active noise cancellation device is invoked to generate an inverse acoustic wave signal corresponding to the predicted audio signal, wherein the inverse acoustic wave signal is an audio signal with the opposite phase to the predicted audio signal; The active noise cancellation device is invoked to output the reverse acoustic wave signal.

9. The mouth-worn device according to claim 8, wherein the processing device is configured to invoke the active noise cancellation device to determine the signal degradation equation of the predicted audio signal; The processing device is used to determine the inverse acoustic wave signal corresponding to the predicted audio signal based on the signal degradation equation. The processing device is used to invoke the anti-phase loudspeaker to output the reverse sound wave signal.

10. The mouth-worn device according to claim 9, wherein the processing device is configured to call a preset speech prediction model to perform audio feature analysis on the predicted audio signal to obtain the predicted audio signal of the wearer after a preset duration; The processing device is used to perform audio degradation processing on the predicted audio signal based on the degradation equation to obtain a degraded audio signal; The processing device is used to perform signal inversion processing on the degraded audio signal to obtain an inverted sound wave signal; The processing device is used to invoke the anti-phase loudspeaker to output the reverse sound wave signal.

11. The mouth-wearing device according to claim 10, wherein the processing device is configured to perform audio degradation processing on the predicted audio signal based on the degradation equation to obtain a degraded audio signal; The processing device is used to perform signal inversion processing on the degraded audio signal to obtain an inverted sound wave signal.

12. The mouth-worn device according to claim 11, wherein the processing device is configured to acquire the audio acoustic features and phoneme sequence of the first audio signal based on a speech prediction model; The processing device is used to generate a predicted phoneme sequence based on the audio acoustic features and the phoneme sequence; The processing device is used to convert the predicted phoneme sequence into an audio signal to obtain a predicted audio signal.

13. The mouth-worn device according to claim 11, wherein the processing device is configured to acquire the audio acoustic features and phoneme sequence of the predicted audio signal based on a speech prediction model; The processing device is used to generate a predicted phoneme sequence based on the audio acoustic features and the phoneme sequence; The processing device is used to convert the predicted phoneme sequence into an audio signal to obtain a predicted audio signal.

14. The mouth-wearing device according to claim 8, wherein the processing device is configured to acquire a second audio signal from the user when the active speaker function of the mouth-wearing device is enabled; The processing device is used to output the second audio signal based on the external speaker.

15. The mouth-worn device according to claim 14, wherein the processing device is configured to acquire preset configuration information of the wearer for the active speaker function; The processing device is used to optimize the second audio signal based on the preset configuration information, and output the optimized second audio signal through the external speaker.

16. The mouth-wearing device according to claim 14, wherein the processing device is configured to, when both the active sound-speaking function and the active sound-canceling function of the mouth-wearing device are enabled, output a reverse sound wave signal of the predicted audio signal based on the reverse speaker, and output a third audio signal of the wearer based on the external speaker.

17. The mouth-worn device according to claim 8, wherein the active noise cancellation device is provided with a first microphone and a reverse speaker, wherein the first microphone is a unidirectional microphone.

18. The mouth-worn device according to claim 8, wherein the outer casing is provided with a second microphone, the second microphone being an omnidirectional microphone.

19. The mouth-wearing device according to claim 8, wherein the sound-absorbing cavity is composed of an inner cavity and an outer cavity, and the space between the inner cavity and the outer cavity is filled with sound-absorbing foam.

20. The mouth-wearing device according to claim 8, wherein a first rubber sealing ring is provided at the edge of the cavity in the area of ​​the user's mouth in the anechoic cavity, and a second rubber sealing ring is provided at the shell in the area of ​​the user's face in the outer shell.

21. The mouth-wearing device according to claim 8, wherein the mouth-wearing device includes a battery, the battery being electrically connected to the processing device, the external speaker, the active noise cancellation device, the first microphone, and the second microphone.

22. An electronic device, the electronic device comprising: Memory, used to store executable program code; A processor for calling and running the executable program code from the memory, causing the electronic device to perform the method as described in any one of claims 1 to 7.

23. A computer-readable storage medium storing a computer program that, when executed, implements the method as described in any one of claims 1 to 7.

24. A computer program product having at least one instruction stored thereon, wherein the at least one instruction, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.

Images