Voice call method, device and readable storage medium based on noise-cancelling mask

Abstract

The present application discloses a voice call method, device and readable storage medium based on a silencer mask, which relates to the field of data processing technology. The voice call based on the silencer mask includes: if the silencer mask detects that the communication module establishes an external call, the user's voice filtered by the sound insulation cotton is collected through a microphone; intelligent voice data is obtained based on the voiceprint feature reconstruction of the user's voice; and the intelligent voice data is sent to the communication module, so that the communication module makes an external call based on the intelligent voice data. The present application achieves the technical effect of protecting the user's call privacy.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a voice call method, device and readable storage medium based on a noise-cancelling mask. Background Art

[0002] When making calls in public places, the content of the call can easily be heard by people around you. With traditional methods of making calls, such as using a mobile phone to make calls directly, people around you may inadvertently hear the content of the call. The leakage of the content of the call will pose a threat to the user's safety. For example, sensitive information of the user involved in the call may be used by criminals, posing a risk of privacy leakage.

[0003] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The main purpose of this application is to provide a voice call method, device and readable storage medium based on a noise-cancelling mask, aiming to solve the technical problem of how to protect the user's call privacy.

[0005] To achieve the above purpose, the present application proposes a voice call method based on a noise-cancelling mask, wherein the noise-cancelling mask is provided with sound-proof cotton to filter the original voice of the user, and the noise-cancelling mask is connected to a communication module. The voice call method based on the noise-cancelling mask includes: If the silencer mask detects that the communication module has established an external call, the user's voice filtered by the sound insulation cotton is collected through the microphone; Reconstructing the voiceprint features of the user's voice to obtain intelligent voice data; The intelligent voice data is sent to the communication module, so that the communication module makes an external call based on the intelligent voice data.

[0006] In one embodiment, if the noise-cancelling mask detects that the communication module establishes an external call, before the step of collecting the user's voice filtered by the sound-proof cotton through the microphone, it includes: According to at least one target terminal selected by a user, acquiring a communication address of the target terminal; generating an external call request based on the communication address of the communication module; According to the communication address of the target terminal, the external call request is sent to each of the target terminals through the communication module to establish the external call between the communication module and the target terminal.

[0007] In one embodiment, the step of reconstructing the intelligent voice data based on the voiceprint feature of the user's voice includes: Filtering the user's voice and normalizing the volume of the user's voice; Extracting voiceprint features of the user's voice to reflect the acoustic characteristics of the user's voice; The voiceprint features are input into an artificial intelligence model, the voiceprint features are reconstructed, and the intelligent voice data is generated.

[0008] In one embodiment, the step of extracting the voiceprint features of the user's voice to reflect the acoustic characteristics of the user's voice includes: Extracting preliminary voiceprint features of the user's voice; Obtaining the sound insulation frequency range of the sound insulation cotton; According to the sound insulation frequency range of the sound insulation cotton, the preliminary voiceprint feature is compensated and processed to obtain the voiceprint feature.

[0009] In one embodiment, the step of compensating the preliminary voiceprint feature according to the sound insulation feature of the sound insulation cotton to obtain the voiceprint feature includes: Determine user characteristics based on user information bound to the noise-cancelling mask; The target voiceprint in the preliminary voiceprint feature is extracted according to the sound insulation frequency range, and the target voiceprint is enhanced by the user feature to obtain the voiceprint feature.

[0010] In one embodiment, the step of inputting the voiceprint feature into an artificial intelligence model, reconstructing the voiceprint feature, and generating the intelligent voice data includes: Inputting the voiceprint feature into the artificial intelligence model; By means of the artificial intelligence model, the voiceprint features are encoded into low-dimensional voiceprint features, key information of the low-dimensional voiceprint features is enhanced and noise is removed; Decode the low-dimensional voiceprint features, reconstruct the voiceprint features, and generate the intelligent voice data.

[0011] In one embodiment, before the step of inputting the voiceprint feature into an artificial intelligence model, reconstructing the voiceprint feature, and generating the intelligent voice data, the step includes: Collecting voiceprint feature sample data from various sources and preprocessing the voiceprint feature sample data; Dividing the voiceprint feature sample data into a training set, a validation set and a test set, and inputting the training set into an initial model for training; Adjusting the parameters of the initial model through an optimization algorithm to minimize the loss function and determine the artificial intelligence model; The artificial intelligence model is evaluated based on the validation set and the test set, and the artificial intelligence model is adjusted according to the evaluation results.

[0012] In one embodiment, the step of sending the intelligent voice data to the communication module so that the communication module makes an external call based on the intelligent voice data includes: Sending the intelligent voice data to the communication module; Encoding the intelligent voice data through the communication module and sending it to the call end corresponding to the external call; Receive the intelligent voice data sent by the call end that is decoded by the communication model, and play the intelligent voice data in real time.

[0013] In addition, to achieve the above-mentioned purpose, the present application also proposes a voice call device based on a noise-cancelling mask, the device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the voice call method based on a noise-cancelling mask as described above.

[0014] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the voice call method based on the noise-cancelling mask as described above are implemented.

[0015] The present application provides a voice call method based on a noise-cancelling mask. The present application detects that a communication module has established an external call, collects the user's voice after filtering with soundproof cotton, reconstructs the user's voiceprint features to obtain intelligent voice data, and sends the intelligent voice data to the communication module, so that the communication module can make an external call based on the intelligent voice data. By filtering with soundproof cotton and reconstructing the user's voiceprint features, the possibility of people around hearing the content of the call is reduced, and the user's call privacy is protected. By reconstructing the intelligent voice data using voiceprint features, the clarity of the call is improved. The present application achieves the technical effect of protecting the user's call privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0018] Figure 1 A flow chart of the first embodiment of the voice call method based on a noise-cancelling mask provided in the present application; Figure 2 A flow chart of the second embodiment of the voice call method based on a noise-cancelling mask provided in this application; Figure 3 A flow chart of the third embodiment of the voice call method based on a noise-cancelling mask provided in the present application; Figure 4 A flow chart of the fourth embodiment of the voice call method based on a noise-cancelling mask provided in the present application; Figure 5 A flowchart of the fifth embodiment of the voice call method based on a noise-cancelling mask provided in the present application; Figure 6 A flowchart diagram of the sixth embodiment of the voice call method based on a noise-cancelling mask provided in the present application; Figure 7 A schematic diagram of the structure of a noise-cancelling mask involved in the voice call method based on a noise-cancelling mask in an embodiment of the present application; Figure 8 This is a schematic diagram of the internal structure of the noise-cancelling mask without the rear shell involved in the voice call method based on the noise-cancelling mask in the embodiment of the present application; Fig. 9 This is a schematic diagram of the device structure of the hardware operating environment involved in the voice call method based on a noise-cancelling mask in an embodiment of the present application.

[0019] Reference numerals: 21. Silencer assembly; 211. First audio connection hole; 212. Silencer shell exhaust hole; 213. Strap hole; 214. Silencer shell exhaust pipe; 215. Silencer shell; 216. Microphone fixing tube; 217. Connector; 218. Front shell; 219. Back shell; 22. Skin-friendly sound insulation pad; 23. Silencer box.

[0020] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0021] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0022] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0023] The main solutions of the embodiments of this application are: At present, when making calls in public places, the content of the call can easily be heard by people around. With traditional methods of making calls, such as using mobile phones directly to make calls, people around may inadvertently hear the content of the call. The leakage of call content will pose a threat to the user's safety. For example, sensitive information of the user involved in the call may be used by criminals, posing a risk of privacy leakage.

[0024] The present application provides a voice call method based on a noise-cancelling mask. The present application detects that a communication module has established an external call, collects the user's voice after filtering with soundproof cotton, reconstructs the user's voiceprint features to obtain intelligent voice data, and sends the intelligent voice data to the communication module, so that the communication module can make an external call based on the intelligent voice data. By filtering with soundproof cotton and reconstructing the user's voiceprint features, the possibility of people around hearing the content of the call is reduced, and the user's call privacy is protected. By reconstructing the intelligent voice data using voiceprint features, the clarity of the call is improved. The present application achieves the technical effect of protecting the user's call privacy.

[0025] It should be noted that the execution subject of this embodiment can be a silencer mask, or a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or a control device of a silencer mask capable of realizing the above functions, etc., and this embodiment does not specifically limit this. The following takes the silencer mask as the execution subject as an example to illustrate this embodiment and the following embodiments.

[0026] Embodiment 1 Based on this, the present application proposes a voice call method based on a noise-cancelling mask in the first embodiment, please refer to Figure 1 , the voice call method based on the noise-cancelling mask includes: Step S10: If the noise-cancelling mask detects that the communication module has established an external call, the user's voice filtered by the noise-cancelling cotton is collected through the microphone.

[0027] The original voice is processed for noise reduction through the sound insulation cotton built into the silencer mask, thereby reducing the spread of the original sound in the external environment to protect the user's privacy and prevent the content of the call from being heard by people around when making voice calls in public places. At the same time, it reduces the interference of environmental noise on call quality and improves the user's calling experience.

[0028] In this embodiment, the communication module is connected to the noise-cancelling mask and is responsible for realizing the communication function between devices, including the communication chip of the mobile phone and the communication interface in the noise-cancelling mask. External calls are voice call connections established with external devices through the communication module, and external devices include mobile phones, landlines, and other noise-cancelling masks. Sound-proof cotton is a material set in the noise-cancelling mask to filter the user's original voice, and has good sound absorption and sound insulation properties.

[0029] Optionally, after collecting the user voice filtered by the sound insulation cotton, the collected user voice is analyzed in real time, and an active noise reduction algorithm is used to generate a sound signal with a phase opposite to the user voice to offset the user voice, thereby further filtering the user voice.

[0030] Exemplarily, based on the original sound captured by the microphone, an algorithm is used to calculate and generate a sound wave signal with a phase opposite to the captured original sound, which is played through the external speaker of the mask, meets and cancels out the user's original sound signal to achieve a noise reduction effect.

[0031] In this embodiment, the noise in the external environment is isolated by sound insulation cotton, so as to reduce the interference of environmental noise on the user's voice, improve the clarity of the voice, and improve the voice quality.

[0032] Exemplarily, the sound insulation cotton absorbs or reflects the noise in the external environment, thereby reducing the interference of the noise on the original voice received by the microphone.

[0033] Step S20, reconstructing the voiceprint features of the user's voice to obtain intelligent voice data.

[0034] Through audio processing technology and artificial intelligence algorithms, voiceprint features are extracted from the original user voice, and clearer intelligent voice data is reconstructed based on the voiceprint features to improve the clarity of voice calls, allowing users to maintain high-quality language communication in noisy environments.

[0035] In this embodiment, voiceprint features refer to the unique acoustic characteristics generated by the user during the pronunciation process, including pitch, timbre, speaking speed, rhythm, etc. Intelligent voice data is voice data processed by algorithms. On the basis of retaining the basic information of the original voice, the clarity of the voice is improved by means of enhancement, denoising, reconstruction, etc. Reconstruction is to use algorithms to recombine and process user voice data to generate new intelligent voice data. The reconstruction process involves denoising, enhancement, encoding, etc.

[0036] As an optional implementation, the user's voice is feature extracted through a machine learning algorithm to identify voiceprint features, and based on the extracted voiceprint features, intelligent voice data is reconstructed using speech synthesis technology or a speech enhancement algorithm.

[0037] Step S30, sending the intelligent voice data to the communication module, so that the communication module makes an external call based on the intelligent voice data.

[0038] In the communication session, the communication module realizes the real-time transmission of intelligent voice data and ensures that users of each terminal in the external call can hear these intelligent voices clearly.

[0039] As an optional implementation, the intelligent voice data is directly transmitted to the communication module of the noise-cancelling mask through the communication interface inside the noise-cancelling mask. After receiving the intelligent voice data, the communication module encodes and processes the intelligent voice data and sends the encoded intelligent voice data to the other party through wireless communication technology. Among them, encoding includes steps such as data compression and modulation.

[0040] As another optional implementation, the generated intelligent voice data is transmitted to the communication module of the mobile phone paired with the silencer mask through the communication interface inside the silencer mask, encoded and processed by the communication module on the mobile phone, and the encoded intelligent voice data is sent to the other party through the wireless network established by the mobile phone.

[0041] Optionally, during transmission, the intelligent voice data is encrypted to prevent eavesdropping or disclosure by unauthorized third parties.

[0042] This embodiment provides a voice call method based on a noise-cancelling mask. This embodiment first detects that an external call is established by a communication module, collects the user's voice after filtering with soundproof cotton, reconstructs the user's voiceprint features to obtain intelligent voice data, and sends the intelligent voice data to the communication module, so that the communication module makes an external call based on the intelligent voice data. By filtering with soundproof cotton and reconstructing the user's voiceprint features, the possibility of people around hearing the call content is reduced, the user's call privacy is protected, and the clarity of the call is improved by reconstructing the intelligent voice data using voiceprint features.

[0043] Based on the first embodiment, the second embodiment of the present application proposes a voice call method based on a noise-cancelling mask, referring to Figure 2 , before step S10, including: Step S50: acquiring the communication address of the target terminal according to at least one target terminal selected by the user.

[0044] According to the user's selection, at least one target terminal is determined, and the communication addresses of these target terminals are obtained. After the communication addresses are obtained, a call request is generated so as to send the request to the target terminal, thereby starting the call process.

[0045] In this embodiment, the target terminal is one or more target terminals selected by the user from the list of available devices, including multiple selection methods such as using an application on the mobile phone connected to the noise-cancelling mask, buttons on the noise-cancelling mask device, etc. The target terminal is the device that the user wants to communicate with, including a mobile phone, tablet computer, personal computer, landline phone, noise-cancelling mask device, etc. The communication address is an address used to uniquely identify a device in a network or other communication network, including an IP address, MAC address, telephone number, etc.

[0046] Optionally, there is a database containing all silencer mask devices and their communication addresses in the control system of the silencer mask, and the communication address of the target terminal is obtained by querying the database.

[0047] As an optional implementation, the communication address of the target terminal is obtained according to the target terminal selected by the user on the mobile phone to which the silencer mask is paired.

[0048] As another optional implementation, the communication address of the target terminal is obtained according to the target terminal selected by the user from the terminal device list displayed on the display screen of the noise-cancelling mask.

[0049] Exemplarily, the communication addresses of all the noise-cancelling masks that meet the permissions in the noise-cancelling mask database are filtered according to the user's permissions, and then the filtered noise-cancelling masks are queried according to the user's request to determine the communication address of the target terminal.

[0050] Step S60: generating an external call request based on the communication address of the communication module.

[0051] In this embodiment, the communication address is a unique identifier of the communication module that initiates the call request in the network or communication system. The call request is a request signal sent by the call initiator to the call recipient through a certain communication method to establish a communication session. The communication method includes telephone, Internet phone, instant messaging software, etc.

[0052] Exemplarily, the communication address of the communication module is obtained as the source address to generate an external call request.

[0053] Step S70: sending the external call request to each of the target terminals through the communication module according to the communication address of the target terminal, and establishing the external call between the communication module and the target terminal.

[0054] The generated call request is sent to each target terminal specified by the user through a network connection, ensuring that the call request can be accurately transmitted to the target terminal, thereby establishing a communication session.

[0055] As an optional implementation, when the acquired communication address is the IP (Internet Protocol) address of the target terminal, a network connection with the target terminal is established through the TCP / IP (Transmission Control Protocol / InternetProtocol) protocol stack, and the call request is encapsulated into a TCP (Transmission Control Protocol) or UDP (User DatagramProtocol) data packet and sent to the IP address of the target terminal.

[0056] As another optional implementation, the acquired communication address is the telephone number of a mobile phone paired with the target terminal, and the telephone number of the target terminal is dialed through a Public Switched Telephone Network (PSTN) or a mobile communication network.

[0057] As another optional implementation, the acquired communication address is the Bluetooth address or other short-range wireless communication identifier of the target terminal, a connection is established with the target terminal through the Bluetooth protocol or other short-range wireless communication protocol, and a call request is sent through the connection.

[0058] As an example of this embodiment, the user presets button rules in the system of the silencer mask, including rules for the object of initiating a call request by touching a button. The user initiates a call request through the button on the silencer mask. Based on the user's call request, the silencer mask device nearby is searched through wireless communication technology, and the target silencer mask is determined according to the preset rules to establish a communication session.

[0059] Exemplarily, a user initiates a call request by sending a voice command to the noise-cancelling mask, the user's voice command is recognized, the corresponding target noise-cancelling mask is determined, and a communication session is established.

[0060] Exemplarily, the silencer mask is paired and connected through a mobile phone or other communication device to establish a stable silencer mask communication link. The user dials a phone number through the mobile phone or uses a third-party calling application to initiate a call request. The mobile phone establishes a mobile phone communication link according to the call object corresponding to the request, bridges the silencer mask communication link and the mobile phone communication link, and establishes a communication session.

[0061] As another optional implementation of this embodiment, a call request is received through a communication module, and a communication session is established between requesting terminals corresponding to the call request.

[0062] As an optional implementation, when both users of the call are using noise-cancelling masks with a call function, the call request is received through the built-in communication module and a communication session is directly established.

[0063] Exemplarily, pairing and connection are performed through Bluetooth, Wi-Fi Direct or other short-range communication technologies. After the connection is established, one user can initiate a voice call request through the microphone on the noise-cancelling mask or initiate a call request through the buttons of the noise-cancelling mask, and other users can answer the call through the speakers or headphones on the mask.

[0064] As another optional implementation, when the noise-cancelling mask used by the calling user does not have an independent calling function, a communication session is established through a terminal device with communication function such as a mobile phone as an intermediary.

[0065] Exemplarily, the noise-cancelling mask is paired and connected with a mobile phone, a call request is initiated or received through the mobile phone, the sound received by the microphone of the noise-cancelling mask is transmitted to the mobile phone, and the sound received by the mobile phone is transmitted to the speaker of the noise-cancelling mask to establish a communication session.

[0066] Exemplarily, a call request from a requesting terminal is received and parsed, and based on the parsed information, a user information database is queried to determine the communication address of the requesting terminal, and a communication session with the requesting terminal is established.

[0067] The present embodiment provides a voice call method based on a noise-cancelling mask. The present embodiment first obtains the communication address of the target terminal to ensure that the call request is accurately sent to the correct target terminal. The call request is sent to the target terminal in real time through the network, which reduces the waiting time and improves the efficiency of establishing a communication session. After the initiating end receives the confirmation response, a communication session is established between the initiating end and the target terminal. The immediate return of the confirmation response and the rapid establishment of the communication session further improve the communication efficiency.

[0068] Based on Example 1, Example 3 of the present application proposes a voice call method based on a noise-cancelling mask, referring to Figure 3 , step S20 comprises: Step S21, filtering the user voice and normalizing the volume of the user voice.

[0069] Filtering can remove noise and interference from speech signals to make speech clearer, and normalization helps eliminate volume differences and keeps the original speech at different volume levels consistent in amplitude, thereby improving the stability of speech processing.

[0070] In this embodiment, filtering is a signal processing technology that removes unnecessary frequency components in the signal through algorithms or devices, retains or enhances the required frequency components, and is used to remove interference factors such as background noise and echo to improve the clarity of speech. Normalization processing is to adjust speech signals of different volume levels to have the same maximum or average amplitude. The volume level is the strength of the sound, usually in decibels, and different volume levels correspond to different sound amplitudes.

[0071] As an optional implementation, the user voice signal is filtered by a filter, and the volume level of the user voice signal is measured, including parameters such as the maximum amplitude, average amplitude or peak value. According to the measured volume level and the determined target amplitude range, a normalization algorithm is applied to perform normalization processing. Among them, the target amplitude range is a normalized target amplitude range determined according to the requirements of the application scenario.

[0072] Exemplarily, the amplitude of the user voice signal is linearly mapped to the target amplitude range through a linear normalization algorithm. If the amplitude range of the original signal is "[A_min, A_max]", and the target amplitude range is "[B_min, B_max]", each amplitude value x of the original signal can be mapped to the target amplitude value y through a linear transformation, that is, "y=((x-A_min) / (A_max-A_min))*(B_max-B_min) +B_min".

[0073] Exemplarily, the amplitude value of the user voice signal is converted into a logarithmic scale through a logarithmic normalization algorithm, and then normalized.

[0074] Step S22: extracting the voiceprint features of the user's voice to reflect the acoustic characteristics of the user's voice.

[0075] In this embodiment, the voiceprint is a sound wave spectrum that carries speech information displayed by an electroacoustic instrument. It is a biological feature composed of more than 100 characteristic dimensions such as wavelength, frequency, and intensity. Voiceprint features are extracted from voiceprint information and can be used to distinguish different individuals or express key information of voice attributes.

[0076] As an optional implementation, a signal processing algorithm is used to analyze the user's voice to extract voiceprint features that can reflect its acoustic characteristics. The signal processing algorithm includes Short-Time Fourier Transform (STFT), Mel-Frequency Cepstral Coefficients (MFCC), Linear Predictive Coding (LPC), etc.

[0077] Step S23, inputting the voiceprint features into an artificial intelligence model, reconstructing the voiceprint features, and generating the intelligent voice data.

[0078] In this embodiment, the artificial intelligence model is a model trained using an artificial intelligence algorithm, which can process, analyze or recognize features in a speech signal and output the processed results.

[0079] Exemplarily, the extracted speech features are input into an artificial intelligence model, the input speech features are processed by the artificial intelligence model, and the relationship between the features is analyzed to reconstruct the voiceprint features to generate enhanced intelligent speech.

[0080] Optionally, step S23 includes: Step A10, inputting the voiceprint features into the artificial intelligence model.

[0081] Step A20, encoding the voiceprint features into low-dimensional voiceprint features through the artificial intelligence model, enhancing the key information of the low-dimensional voiceprint features and removing noise.

[0082] In this embodiment, the low-dimensional voiceprint feature refers to a low-dimensional representation obtained by converting a high-dimensional voiceprint feature into a low-dimensional representation through a dimensionality reduction algorithm. The low-dimensional representation can retain key information in the original feature while removing redundancy and noise.

[0083] Exemplarily, through the artificial intelligence model, the encoding layer is used to convert high-dimensional voiceprint features into low-dimensional representations, the noise in the low-dimensional representation is removed by the noise reduction algorithm, and the key information in the low-dimensional representation is enhanced by the feature selection algorithm.

[0084] Step A30, decoding the low-dimensional voiceprint features, reconstructing the voiceprint features, and generating the intelligent voice data.

[0085] Exemplarily, through the artificial intelligence model, in the decoding layer, the low-dimensional voiceprint features are converted back to high-dimensional representations, and reconstructed voiceprint features similar to the original voiceprint features are generated through a reconstruction algorithm, which are further filtered and processed to obtain intelligent voice data.

[0086] The present embodiment provides a voice call method based on a noise-cancelling mask. The present embodiment firstly performs filtering and normalization processing to effectively remove noise and interference in the user's voice and improve the clarity of the voice. By extracting voiceprint features that can reflect the user's acoustic characteristics and combining them with an artificial intelligence model for processing and analysis, the accuracy of the voiceprint features can be enhanced to generate accurate and clear intelligent voice data.

[0087] Based on the first and third embodiments, the fourth embodiment of the present application proposes a voice call method based on a noise-cancelling mask, referring to Figure 4 , step S22 comprises: Step A40: extracting preliminary voiceprint features of the user's speech.

[0088] In this embodiment, the preliminary voiceprint feature refers to a feature extracted from the user's voice signal that can preliminarily reflect the acoustic characteristics of the user's voice.

[0089] Exemplarily, a feature extraction algorithm is used to extract features from the user's voice signal to obtain preliminary voiceprint features.

[0090] Step A50, obtaining the sound insulation frequency range of the sound insulation cotton.

[0091] In this embodiment, the sound insulation frequency range refers to the frequency range in which the sound insulation cotton can effectively isolate sound, which is related to factors such as the material, structure, and density of the sound insulation cotton.

[0092] As an optional implementation, the sound insulation frequency range is determined according to the product manual or technical documentation of the sound insulation cotton.

[0093] As another optional implementation, professional equipment is used to perform a frequency response test on the sound insulation cotton to determine the sound insulation frequency range.

[0094] Step A60, compensating and processing the preliminary voiceprint feature according to the sound insulation frequency range of the sound insulation cotton to obtain the voiceprint feature.

[0095] Compensation processing is performed on the initially extracted user voice print features to eliminate or reduce the impact of the sound insulation cotton on the voice print features. As a sound insulation material, the sound insulation cotton absorbs, reflects or transmits sound waves within a specific frequency range, which may cause certain frequency components in the user's voice to be weakened or enhanced. By performing compensation processing based on the sound insulation frequency range of the sound insulation cotton, these affected frequency components can be restored or enhanced, thereby obtaining a more accurate and complete voice print feature.

[0096] In this embodiment, the compensation processing refers to correcting and enhancing the preliminary voiceprint characteristics according to the sound insulation frequency range of the sound insulation cotton to eliminate the influence of the sound insulation cotton on the voiceprint characteristics.

[0097] Exemplarily, the sound insulation frequency range of the sound insulation cotton is analyzed to determine the influence range of the sound insulation frequency range on the preliminary voiceprint features, and the compensation algorithm is used to correct and enhance the preliminary voiceprint features.

[0098] Optionally, step A60 includes: Step A61, determining user characteristics according to the user information bound to the noise-cancelling mask.

[0099] In this embodiment, the user information is data related to the user bound to the noise-cancelling mask, including the user's identity information, voice characteristics, usage habits, etc.

[0100] Step A62, extracting a target voiceprint from the preliminary voiceprint feature according to the sound insulation frequency range, and strengthening the target voiceprint through the user feature to obtain the voiceprint feature.

[0101] When the user wears a noise-cancelling mask, the target voiceprint is accurately identified and extracted from the initially extracted voiceprint features, and these target voiceprints are enhanced using the user's personal characteristics to obtain clearer voiceprint features.

[0102] Exemplarily, based on the sound insulation frequency range, the preliminary voiceprint features are analyzed in the frequency domain, the target voiceprint that is less affected by the silencer mask is extracted from the preliminary voiceprint features, the extracted user features are used to enhance the target voiceprint, and the enhanced target voiceprint is used as the final voiceprint feature.

[0103] The present embodiment provides a voice call method based on a noise-cancelling mask. The present embodiment first extracts preliminary voiceprint features to preliminarily reflect the acoustic characteristics of the user's voice. By obtaining the sound insulation frequency range of the sound insulation cotton and performing compensation processing accordingly, the influence of the sound insulation cotton on the voiceprint features can be eliminated or reduced, thereby improving the accuracy of the voiceprint features.

[0104] Based on the first and third embodiments, the fifth embodiment of the present application proposes a voice call method based on a noise-cancelling mask, referring to Figure 5 , before step S23, including: Step S24, collecting voiceprint feature sample data from various sources and preprocessing the voiceprint feature sample data.

[0105] In this embodiment, voiceprint feature sample data refers to voiceprint features collected from various sources, including features of different speakers, and is used to train and evaluate artificial intelligence models. Preprocessing is a series of operations performed on voiceprint feature sample data, including denoising, filtering, normalization, feature extraction, etc., to improve data quality and make the voiceprint feature sample data more suitable for model training.

[0106] As an optional implementation, according to the collection requirements of voiceprint feature sample data, sample data is collected from various sources, the sample data is preprocessed using signal processing technology, and voiceprint features are extracted for subsequent model training. The collection requirements include the type of voice samples and the number of collected samples. The types of voice samples include dialects, accents, speaking speed, volume, sound quality and other types.

[0107] Exemplarily, sample data is collected from a public speech data set, invalid or damaged sample data is removed, useful voiceprint features are extracted from the sample data as voiceprint feature sample data, and attributes such as amplitude and volume of the voiceprint feature sample data are scaled to a standard range.

[0108] Step S25, dividing the voiceprint feature sample data into a training set, a validation set and a test set, and inputting the training set into the initial model for training.

[0109] In this embodiment, the training set is the data divided from the voiceprint feature sample data for training the artificial intelligence model. The validation set is a data set used to evaluate the model performance during the model training process. The test set is a data set used to evaluate the generalization ability of the model after the model training is completed. The initial model refers to an untrained artificial intelligence model.

[0110] For example, an initial model is selected according to task requirements, and a loss function that reflects the performance of the model is designed, which can accurately measure the performance of the model in enhancing useful components and suppressing noise. The training set data is input into the initial model for training.

[0111] Step S26, adjusting the parameters of the initial model through an optimization algorithm to minimize the loss function and determine the artificial intelligence model.

[0112] In this embodiment, the optimization algorithm is an algorithm for adjusting model parameters to minimize the loss function, including gradient descent, stochastic gradient descent, Adam (Adaptive Moment Estimation) and other algorithms. The loss function is an indicator that measures the difference between the model prediction results and the actual results, and the model parameters are optimized by minimizing the loss function.

[0113] As an optional implementation, hyperparameters such as learning rate, momentum, and weight decay are adjusted according to the selected optimization algorithm.

[0114] Exemplarily, a training set is used, a loss function value is calculated by forward propagation, a gradient is calculated by a back-propagation algorithm, and model parameters are updated using an optimization algorithm, and training is repeated until the loss function value converges.

[0115] Step S27, evaluating the artificial intelligence model based on the verification set and the test set, and adjusting the artificial intelligence model according to the evaluation results.

[0116] As an optional implementation, during the model training process, the validation set is used regularly to evaluate the performance of the model, and the loss function value, enhancement effect, noise suppression ability and other indicators on the validation set are calculated. According to the evaluation results, the parameters of the model are adjusted to optimize the performance of the model. After the model training is completed, the test set is used for the final evaluation, and the loss function value, enhancement effect, noise suppression ability and other indicators on the test set are calculated and compared with the results on the validation set. According to the evaluation results on the validation set and the test set, the model is adjusted. Among them, the adjustment includes changing the model architecture, adjusting the hyperparameters, etc.

[0117] Alternatively, if the performance on the test set is significantly different from that on the validation set, it indicates that the model has overfitting or underfitting problems, and the structure and parameters of the model need to be further checked.

[0118] The present embodiment provides a voice call method based on a noise-cancelling mask. The present embodiment first ensures the diversity of training data by collecting voiceprint feature sample data from different sources, and divides the voice sample data into a training set, a validation set, and a test set, thereby ensuring the independence of model training, validation, and testing, and facilitating objective performance evaluation during the model training process to avoid overfitting or underfitting. The model parameters are adjusted through an optimization algorithm, and the model is continuously adjusted and optimized based on the evaluation results of the validation set and the test set, thereby forming an iterative optimization cycle, which continuously improves the model performance and improves the accuracy of the model.

[0119] Based on Example 1, Example 6 of the present application proposes a voice call method based on a noise-cancelling mask, referring to Figure 6 , after step S30, comprising: Step S80, sending the intelligent voice data to the communication module.

[0120] Exemplarily, the generated intelligent voice data is sent to the communication module through the internal interface of the noise-cancelling mask.

[0121] Step S90, encoding the intelligent voice data through the communication module and sending it to the call end corresponding to the external call.

[0122] In this embodiment, encoding is a technology that converts intelligent voice data into a compressed format, aiming to reduce the data size while maintaining the audio quality as much as possible to facilitate the transmission of voice data. The calling end is the other end device that makes a voice call with the noise cancelling mask, including mobile phones, landlines, other noise cancelling mask devices, etc.

[0123] Exemplarily, an audio coding algorithm suitable for real-time communication is selected, intelligent voice is input into the audio coding algorithm, parameters of the audio coding algorithm are adjusted according to the bandwidth requirements and audio quality requirements of the communication session, and the encoded voice data is sent to the call end corresponding to the external call.

[0124] Optionally, the selection of network transmission protocol and parameter settings are determined based on factors such as the real-time nature of the communication session, delay requirements, bandwidth limitations, etc.

[0125] Step S100, receiving the intelligent voice data sent by the call end and decoded by the communication model, and playing the intelligent voice data in real time.

[0126] In this embodiment, decoding is the process by which the receiving end restores the received encoded data to the original intelligent voice data, which is the reverse process of encoding.

[0127] Exemplarily, the encoded intelligent voice data sent by the call end through the network transmission protocol is received, and the encoded intelligent voice data is decoded into the original intelligent voice data and played in real time according to the inverse process of the encoding algorithm used by the encoded intelligent voice data.

[0128] Optionally, after additional audio processing is performed on the decoded intelligent voice data, the intelligent voice data is played in real time, and the audio processing includes volume adjustment, equalizer, etc.

[0129] The present embodiment provides a voice call method based on a noise-cancelling mask. The present embodiment first sends the intelligent voice data to the communication module, and uses the communication module to encode and send it, so as to ensure that the intelligent voice data can be transmitted to the call end corresponding to the external call in real time and efficiently. The communication module encodes the intelligent voice data, and can optimize the format and rate of data transmission, thereby improving the clarity of communication. At the same time, by receiving and decoding the intelligent voice data from the call end in real time, the quality and efficiency of communication are improved.

[0130] Based on Example 1, Example 7 of the present application proposes a voice call method based on a noise-cancelling mask, referring to Figure 7 and Figure 8 , Figure 7 and Figure 8 A schematic diagram of a noise-cancelling mask involved in a voice call method based on a noise-cancelling mask is provided, specifically: In this embodiment, the silencing device includes a silencing assembly 21, which includes a front shell 218 and a rear shell 219. The front shell 218 and the rear shell 219 are combined to form a mask-shaped structure. A strap hole 213 is provided at the end of the front shell 218, and an adjustable strap is passed through the strap hole 213 and then tied to the silencing device.

[0131] A recessed muffler shell 215 is arranged in the front shell 218, and the bottom and inner wall of the inner cavity of the muffler shell 215 are covered with sound insulation cotton. When the sound passes through the sound insulation cotton, the sound insulation cotton is used to muffle the sound and reduce the vibration caused by the sound. A microphone fixing cylinder 216 is arranged inside the muffler shell 215, and the microphone is installed in the microphone fixing cylinder 216. A muffler box 23 is arranged in the interlayer between the front shell 218 and the rear shell 219, and the muffler shell exhaust pipe 214 connects the muffler shell 215 and the muffler box 23.

[0132] A connector 217 is arranged on the surface of the front shell 218 around the periphery of the entrance of the silencer shell 215, and a skin-friendly sound-isolating pad 22 is pasted on the connector 217. When the user wears the silencer device, the skin-friendly sound-isolating pad 22 is aligned to the outer peripheral side of the user's lips for sealing to reduce sound leakage.

[0133] The silencing device is provided with a first audio connection hole 211, which is connected to the communication module via an audio connection line.

[0134] The user's input voice is first silenced by the sound insulation cotton covered by the silencer shell 215. The airflow of the input voice is then transmitted to the silencer box 23 through the silencer shell exhaust pipe 214.

[0135] Furthermore, in the silencer box 23, at least two boxes with two parallel plates are provided, and the airflow of the input voice enters one of the boxes, propagates in the cavity of the box and repeatedly hits the two parallel plates, vibrates back and forth, and then transmits the sound to the adjacent box. Through the principle of mutual rebound and cancellation of parallel silencer sound waves in the closed silencer cavity, the sound wave energy is weakened, thereby further reducing the speech volume and achieving silencer again.

[0136] The front shell 218 is provided with a silencer shell exhaust hole 212, and the air flow processed by the silencer box 23 is discharged from the silencer shell exhaust hole 212 to ensure air circulation.

[0137] Optionally, the noise reduction cotton in the sound-absorbing shell 215 absorbs the sound waves rebounding from the inner wall of the cavity, ensuring that the sound collected by the microphone is clearer.

[0138] The microphone provided in the silencer shell 215 of the silencer device collects the user's voice after silencer and noise reduction processing, performs voiceprint feature extraction on the collected user's voice, identifies the user's unique voiceprint, and based on the voiceprint feature, uses an artificial intelligence algorithm to enhance the sound waveform, and reconstructs the intelligent voice data.

[0139] Optionally, according to the sound insulation frequency range of the sound insulation cotton, the initially extracted user voice print features are compensated to determine the target voice print that is less affected by the silencer mask. According to the user information bound to the silencer mask, the user features are extracted to strengthen the target voice print, and the strengthened target voice print is used as the final voice print feature, which is then input into the artificial intelligence model to reconstruct the intelligent voice data.

[0140] Optionally, the communication module detects whether an external connection has been established. If no external connection has been established, the communication module waits for the user to connect a mobile phone via a 2.4G signal band and / or connect a Bluetooth headset via Bluetooth.

[0141] The silencing device sends the reconstructed intelligent voice data to the connected communication module through the first audio connection hole 211.

[0142] After receiving the intelligent voice data, the communication module converts it into a recognizable sound signal for external calls.

[0143] Optionally, if all users making external calls wear silencer masks, a multi-person pairing online mode is adopted, and the communication module directly connects to other users wearing silencer masks through the 2.4G signal frequency band to achieve private conversation.

[0144] When the user ends the call, the communication module disconnects.

[0145] The entire process is designed to achieve an efficient, private and intelligent voice call method based on a silencer mask. By integrating silencer technology and intelligent voice processing algorithms, it can significantly reduce the volume of the user's speech while ensuring the clarity of the sound quality, allowing users to communicate effectively while maintaining privacy. Through the combination of physical silencer and intelligent noise reduction technology, it ensures that the user's voice will not be easily heard by the outside world. The multi-person pairing online mode of the silencer mask is directly connected through the 2.4G signal frequency band without relying on external devices, further enhancing the privacy of communication. The application of intelligent voice reconstruction technology makes the transmission of sound signals clearer and improves call quality.

[0146] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the voice call method based on the noise-cancelling mask of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0147] The present application provides a voice call device based on a noise-cancelling mask, and the voice call device based on a noise-cancelling mask includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the voice call method based on the noise-cancelling mask in the above-mentioned embodiment one.

[0148] Reference below Fig. 9 , which shows a schematic diagram of the structure of a voice call device based on a noise-cancelling mask suitable for implementing the embodiment of the present application. The voice call device based on a noise-cancelling mask in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Fig. 9 The voice call device based on the noise-cancelling mask shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0149] like Fig. 9As shown, the voice call device based on the noise cancelling mask may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM: Read Only Memory) 1002 or the program loaded from the storage device 1003 to the random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the voice call device based on the noise cancelling mask are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. The input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 can allow the voice call device based on the noise cancelling mask to communicate with other devices wirelessly or wired to exchange data. Although the figure shows a voice call device based on a noise cancelling mask with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have alternatively.

[0150] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0151] The voice call device based on the silencer mask provided by the present application adopts the voice call method based on the silencer mask in the above embodiment, which can solve the technical problem of how to protect the user's call privacy. Compared with the prior art, the beneficial effects of the voice call device based on the silencer mask provided by the present application are the same as the beneficial effects of the voice call method based on the silencer mask provided by the above embodiment, and the other technical features of the voice call device based on the silencer mask are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0152] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0153] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0154] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the voice call method based on the noise-cancelling mask in the above-mentioned embodiment.

[0155] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0156] The above-mentioned computer-readable storage medium can be included in the voice call device based on the noise-cancelling mask; or it can exist independently without being assembled into the voice call device based on the noise-cancelling mask.

[0157] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the voice call device based on the noise cancelling mask, the voice call device based on the noise cancelling mask can be written in one or more programming languages ​​or a combination thereof to write computer program codes for performing the operations of the present application. The programming languages ​​include object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0158] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0159] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0160] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned voice call method based on a noise-cancelling mask, and can solve the technical problem of how to protect the user's call privacy. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the voice call method based on a noise-cancelling mask provided in the above-mentioned embodiment, and will not be repeated here.

[0161] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A voice call method based on a noise-cancelling mask, characterized in that: The noise-cancelling mask includes noise-isolating cotton and a microphone, the noise-isolating cotton is used to filter the original voice of the user, the microphone is used to collect the filtered original voice, the noise-cancelling mask is connected to the communication module, and the voice call method based on the noise-cancelling mask includes: If the silencer mask detects that the communication module has established an external call, the user's voice filtered by the sound insulation cotton is collected through the microphone; Reconstructing the voiceprint features of the user's voice to obtain intelligent voice data; The intelligent voice data is sent to the communication module, so that the communication module makes an external call based on the intelligent voice data.

2. The voice communication method based on the noise-cancelling mask as claimed in claim 1, characterized in that: Before the step of collecting the user's voice filtered by the sound insulation cotton through the microphone if the silencer mask detects that the communication module establishes an external call, the step includes: According to at least one target terminal selected by a user, acquiring a communication address of the target terminal; generating an external call request based on the communication address of the communication module; According to the communication address of the target terminal, the external call request is sent to each of the target terminals through the communication module to establish the external call between the communication module and the target terminal.

3. The voice communication method based on the noise-cancelling mask as claimed in claim 1, characterized in that: The step of reconstructing the voiceprint feature of the user's voice to obtain intelligent voice data includes: Filtering the user's voice and normalizing the volume of the user's voice; Extracting voiceprint features of the user's voice to reflect the acoustic characteristics of the user's voice; The voiceprint features are input into an artificial intelligence model, the voiceprint features are reconstructed, and the intelligent voice data is generated.

4. The voice communication method based on the noise-cancelling mask as claimed in claim 3, characterized in that: The step of extracting the voiceprint features of the user's voice to reflect the acoustic characteristics of the user's voice includes: Extracting preliminary voiceprint features of the user's voice; Obtaining the sound insulation frequency range of the sound insulation cotton; According to the sound insulation frequency range of the sound insulation cotton, the preliminary voiceprint feature is compensated and processed to obtain the voiceprint feature.

5. The voice communication method based on the noise-cancelling mask as claimed in claim 4, characterized in that: The step of compensating the preliminary voiceprint feature according to the sound insulation feature of the sound insulation cotton to obtain the voiceprint feature comprises: Determine user characteristics based on user information bound to the noise-cancelling mask; The target voiceprint in the preliminary voiceprint feature is extracted according to the sound insulation frequency range, and the target voiceprint is enhanced by the user feature to obtain the voiceprint feature.

6. The voice communication method based on the noise-cancelling mask as claimed in claim 3, characterized in that: The step of inputting the voiceprint features into an artificial intelligence model, reconstructing the voiceprint features, and generating the intelligent voice data comprises: Inputting the voiceprint feature into the artificial intelligence model; By means of the artificial intelligence model, the voiceprint features are encoded into low-dimensional voiceprint features, key information of the low-dimensional voiceprint features is enhanced and noise is removed; Decode the low-dimensional voiceprint features, reconstruct the voiceprint features, and generate the intelligent voice data.

7. The voice communication method based on the noise-cancelling mask as claimed in claim 3, characterized in that: Before the step of inputting the voiceprint feature into an artificial intelligence model, reconstructing the voiceprint feature, and generating the intelligent voice data, the method includes: Collecting voiceprint feature sample data from various sources and preprocessing the voiceprint feature sample data; Dividing the voiceprint feature sample data into a training set, a validation set and a test set, and inputting the training set into an initial model for training; Adjusting the parameters of the initial model through an optimization algorithm to minimize the loss function and determine the artificial intelligence model; The artificial intelligence model is evaluated based on the validation set and the test set, and the artificial intelligence model is adjusted according to the evaluation results.

8. The voice communication method based on the noise-cancelling mask as claimed in claim 1, characterized in that: The step of sending the intelligent voice data to the communication module so that the communication module makes an external call based on the intelligent voice data includes: Sending the intelligent voice data to the communication module; Encoding the intelligent voice data through the communication module and sending it to the call end corresponding to the external call; Receive the intelligent voice data sent by the call end that is decoded by the communication model, and play the intelligent voice data in real time.

9. A voice communication device based on a noise-cancelling mask, characterized in that: The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the voice call method based on a sound-cancelling mask as described in any one of claims 1 to 8.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the voice call method based on a sound-absorbing mask as described in any one of claims 1 to 8 are implemented.

Images