Adjustment method and system for interchangeable earplugs of sleeping headphones based on external sound

Abstract

The present invention relates to the technical field of interchangeable earplugs for sleep headphones, and discloses a method and system for adjusting interchangeable earplugs for sleep headphones based on external sounds, comprising: obtaining data information of external sounds; analyzing the data information and generating characteristic information; using the characteristic information as a trigger signal to adjust the sleep headphones; utilizing a force sensor of the interchangeable earplug to sense the force state of the ear canal on the earplug; and adjusting the pressure of the interchangeable earplug based on the force state and the adjustment result of the sleep headphones. The method can effectively simplify calculations in the case of a single-sided sound source. At the same time, the volume of the headphones and the pressure of the earplug can be adjusted to suit the user's sleeping environment in response to external noise. The method applies neural network technology to analyze time-frequency domain data to improve the accuracy and response speed of sound processing, so that the headphones can more accurately adjust their output according to environmental changes.

Description

Technical Field

[0001] The present invention relates to the technical field of interchangeable earplugs for sleeping headphones, and in particular to an adjustment method and method for the interchangeable earplugs for sleeping headphones based on external sounds. Background Art

[0002] As the pace of life accelerates, sleep problems have gradually become one of the important factors affecting people's health. In addition, with the advancement of technology, people's demand for personalized and intelligent products is also growing.

[0003] Traditional headphones are primarily used for music playback or communication, with few specifically designed for sleep environments. Existing sleep headphones mostly offer only basic noise-blocking capabilities and lack the ability to dynamically adapt to individual sleep environments. Furthermore, traditional headphones often fail to consider long-term wear comfort and ear health, potentially leading to a poor user experience and ear health issues. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the existing adjustment method of the interchangeable earplugs of the sleeping earphones cannot adapt to the external environment in time, and the control cannot adapt to sleep, etc.

[0006] To solve the above technical problems, the present invention provides the following technical solution: a method for adjusting interchangeable earplugs of sleep headphones based on external sounds, comprising:

[0007] Obtain data information of external sounds;

[0008] Analyzing the data information and generating feature information;

[0009] Using the characteristic information as a trigger signal to adjust the sleep headphones;

[0010] The force sensor of the interchangeable earplug is used to sense the force applied by the ear canal to the earplug.

[0011] The pressure of the interchangeable earplug is adjusted according to the force state and the adjustment result of the sleep earphone.

[0012] As a preferred solution of the method for adjusting the interchangeable earplugs of the sleep headphones based on external sounds of the present invention, wherein: the data information of the external sounds obtained includes basic data and advanced data;

[0013] The basic data includes sound intensity data, frequency distribution data, duration and occurrence time of sound events;

[0014] The advanced data includes determining the direction of the sound source based on the basic data obtained from the left and right headphones respectively. The specific steps are as follows:

[0015] Assume that there are sound sensors on both sides of the earphone M1 and M2 respectively, and obtain the time-frequency domain data of M1 and M2;

[0016] In the time-frequency domain image of each sound sensor, peak recognition is performed on each time-frequency domain image segment, and the average value of the peak amplitude in the current time-frequency domain image segment is calculated. Peaks with amplitudes exceeding the average value are screened out and collected as MJ1 and MJ2 respectively;

[0017] MJ1={J1,J2,...,Jn};

[0018] MJ2 = {K1, K2, ..., Km};

[0019] Where Jn represents the nth peak filtered out by M1, Km represents the mth peak filtered out by M2, n represents the maximum peak number filtered out by M1, and m represents the maximum peak number filtered out by M2. Each element in the set represents the size of the peak and its corresponding timestamp.

[0020] Use the time difference between MJ1 and MJ2 to determine the approximate location of the sound source in the earphones;

[0021] Judgment of unilateral sound source: When m=n, according to the number of each element in MJ1 and MJ2, the elements with the same number in MJ1 and MJ2 are matched one-to-one, and the timestamp of each element in MJ1 and MJ2 is obtained; if the timestamp of each element in MJ1 is earlier than the timestamp of the corresponding element in MJ2, the sound source is determined to be on the side of sound sensor M1; if the timestamp of each element in MJ2 is earlier than the timestamp of the corresponding element in MJ1, the sound source is determined to be on the side of sound sensor M2;

[0022] When the judgment of a unilateral sound source is not met, it is judged as a non-unilateral sound source;

[0023] If the sound source is determined to be unilateral, the attenuation of the sound intensity at the sound sensors on both sides is evaluated;

[0024]

[0025] Where i represents the index of the peak number; Indicates the attenuation rate of sound intensity from the sensor close to the sound source to the sensor far away from the sound source; I i大 and I i小 Represents the sound intensity of the two sensors at the i-th peak, and I i大 >I i小 .

[0026] As a preferred embodiment of the method for adjusting the interchangeable earplugs of sleep headphones based on external sound according to the present invention, the data information is analyzed and the characteristic information is generated, which includes: if it is determined that the sound source is unilateral, the time-frequency domain data of the sound sensor on the side close to the sound source are analyzed; if it is determined that the sound source is not unilateral, the time-frequency domain data of the sound sensors on both sides are analyzed respectively;

[0027] The analysis process of time-frequency domain data is completed through neural networks:

[0028] Input layer: Input time-frequency domain image data, which is converted into a format acceptable to the model after preprocessing, denoted as S(t,f);

[0029] TCN layer output:

[0030] TCN output =ReLU(W tcn *S(t,f)+b tcn )

[0031] Among them, W tcn represents the convolution kernel weight in the temporal convolutional network layer, b tcn Represents the bias term in the temporal convolutional network layer;

[0032] LSTM layer output:

[0033] LSTM output =LSTM(TCN output )

[0034] Among them, TCN output Represents the output features at time t after processing by the temporal convolutional network;

[0035] Attention Weighting:

[0036]

[0037] e t represents the attention score at time t:

[0038] e t =v Z tanh(W a LSTM output (t)+b a )

[0039] Where Z represents the transpose; v represents the weight vector of the attention score; W a represents the weight of the attention layer; b a Represents the bias term; τ represents the τth moment predicted from the current moment; T represents the length of the predicted time;

[0040] Weighted summation produces the final feature:

[0041]

[0042] Output prediction:

[0043] [PL, PP] = W0*Attention output +b0

[0044] Where W0 represents the output weight matrix, which is used to linearly transform the feature vector output by the attention mechanism into the final prediction target, i.e., the predicted loudness and pitch; b0 represents the bias vector of the output model; PL represents the predicted loudness; PP represents the predicted pitch;

[0045] PL and PP are used as feature information of a predicted time length T, and the feature information is used to trigger the control of the interchangeable earplugs of the sleeping headset.

[0046] As a preferred embodiment of the method for adjusting the interchangeable earplugs of the sleep earphones based on external sound according to the present invention, the adjustment of the sleep earphones includes obtaining the preferred loudness P set by the user. user , introduces an adaptive adjustment factor based on predicted loudness and pitch, and a dynamic loudness range compressor to optimize volume output;

[0047] Volume adjustment factor:

[0048]

[0049] Among them, α, β, and γ represent adjustment parameters, which are used to balance the sensitivity of loudness adjustment and the periodic effect of pitch; P max Indicates the maximum value of the pitch, used to normalize the pitch value;

[0050] Dynamic loudness range compression:

[0051] V out =V base (1+V factor )·log(1+exp(V factor ))

[0052] Among them, V base Indicates the base volume;

[0053] If it is judged to be a unilateral sound source, V out To control the target, the volume of the earphone close to the sound source is adjusted; and the volume of the earphone far from the sound source is adjusted synchronously according to the attenuation rate. The adjusted volume is expressed as: ρ·V out ;

[0054] If it is determined to be a non-unilateral sound source, the V calculated for the sound on both sides according to their respective feature information is out Perform volume control;

[0055] When adjusting the volume, the headphone volume is adjusted at a preset adjustment speed.

[0056] As a preferred embodiment of the method for adjusting the interchangeable earplugs of sleep headphones based on external sound according to the present invention, wherein: the sensing of the force state of the ear canal on the earplug by using the force sensor of the interchangeable earplug includes embedding a miniature force sensor in the interchangeable earplug to measure the radial force of the ear canal on the earplug;

[0057] If the sound source is determined to be unilateral, the radial force applied to the earplug on the side close to the sound source is controlled based on PP and PL; and the radial force applied to the earplug on the side away from the sound source is simultaneously controlled according to the attenuation rate;

[0058] If it is determined that the sound source is not unilateral, the radial force applied to the earplugs on both sides is controlled separately.

[0059] As a preferred embodiment of the method for adjusting the interchangeable earplugs of sleep headphones based on external sound according to the present invention, the pressure adjustment of the interchangeable earplugs includes controlling the pressure of the interchangeable earplugs on the inside of the ear canal according to the radial force exerted by the inside of the ear canal on the earplugs and the volume of the sleep headphones under the predicted conditions of PP and PL;

[0060] Stress model;

[0061]

[0062] Where P represents the pressure applied by the earplug in the ear canal; It represents the second derivative of pressure with respect to time and is the acceleration of pressure change; represents the spatial derivative term, c is the propagation speed of the sound wave in the ear canal medium; θP represents the nonlinear damping term of the ear plug material, and θ is the damping coefficient; It represents the intensity coefficient of the sound wave source term, which is related to the energy of the sound emitted by the headphones; V represents the headphone volume; ε represents the nonlinear coefficient that adjusts the loudness and volume; PP0 represents the resonant frequency of the ear canal, and δ represents the attenuation coefficient that describes the frequency when it deviates from the resonant frequency;

[0063] Through human-computer interaction, the system is instructed to change the medium. When the user changes the earplug medium, the damping coefficient comparison table of the medium is used to update θ in the nonlinear damping term.

[0064] By solving the pressure model, the radial pressure P exerted by the earplug on the ear canal is obtained, and a control instruction is issued to control the pressure.

[0065] As a preferred embodiment of the method for adjusting the interchangeable earplugs of the sleep earphones based on external sound according to the present invention, the pressure adjustment of the interchangeable earplugs further comprises continuously detecting external sound information on both sides of the left and right ears by a sound sensor of the sleep earphones;

[0066] Based on the continuous detection information of the left and right ears, the volume and pressure of the interchangeable earbuds of the sleep headphones are controlled in real time;

[0067] If the external sound information continuously detected by M1 and the external sound information continuously detected by M2 change at the same time, and the sound characteristics of M1 during the continuous detection are the same as the sound characteristics of M2 before the change, and the sound characteristics of M2 during the continuous detection are the same as the sound characteristics of M1 before the change, then it is determined that the left and right earplugs of the interchangeable sleep headphones have been swapped;

[0068] When the interchangeable earbuds of the sleep headphones are swapped between left and right, the control data of M1 and M2 are replaced so that the left and right ear sides can have the original volume and pressure; after the replacement is completed, continuous detection and real-time control of M1 and M2 continue.

[0069] On the other hand, the sleep headphones have an adjustment system based on external sound with interchangeable ear tips, where:

[0070] The acquisition unit acquires data information about external sounds and uses the force sensor of the interchangeable earplug to sense the force applied by the ear canal to the earplug;

[0071] an analysis unit, for analyzing the data information and generating feature information;

[0072] The control unit uses the characteristic information as a trigger signal to adjust the sleep earphones; and adjusts the radial pressure of the interchangeable earplugs according to the force state and the adjustment result of the sleep earphones.

[0073] A computer device comprises: a memory and a processor; the memory stores a computer program, wherein: when the processor executes the computer program, the steps of any one of the methods of the present invention are implemented.

[0074] A computer-readable storage medium stores a computer program, wherein: when the computer program is executed by a processor, the steps of any one of the methods of the present invention are implemented.

[0075] The beneficial effects of the present invention are as follows: The method for adjusting interchangeable earbuds for sleep headphones based on external sound provided by the present invention can effectively simplify calculations in the case of a single sound source. It can also adjust the headphone volume and earbud pressure to suit the user's sleeping environment in response to external noise. The application of neural network technology to analyze time-frequency domain data improves the accuracy and response speed of sound processing, enabling the headphones to more accurately adjust output based on environmental changes. An interface allows users to adjust headphone settings according to their personal needs, and the system automatically readjusts parameters after earbud media replacement to ensure optimal performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0077] Figure 1 This is an overall flow chart of a method for adjusting interchangeable earplugs of sleep headphones based on external sounds provided by the first embodiment of the present invention. DETAILED DESCRIPTION

[0078] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0079] Example 1

[0080] Reference Figure 1 , as one embodiment of the present invention, provides a method for adjusting interchangeable earplugs of sleep headphones based on external sounds, comprising:

[0081] S1: Obtain data information of external sound.

[0082] The data information obtained from the external sound includes basic data and advanced data.

[0083] The basic data includes sound intensity data, frequency distribution data, duration and occurrence time of sound events.

[0084] The advanced data includes determining the direction of the sound source based on the basic data obtained from the left and right headphones respectively. The specific steps are as follows:

[0085] Assume that there are sound sensors on both sides of the earphone M1 and M2 respectively, and obtain the time-frequency domain data of M1 and M2.

[0086] In the time-frequency domain image of each sound sensor, peak recognition is performed on each segment of the time-frequency domain image, and the average value of the peak amplitude in the current segment of the time-frequency domain image is calculated. Peaks with amplitudes exceeding the average value are screened out and collected as MJ1 and MJ2 respectively.

[0087] MJ1={J1,J2,...,Jn};

[0088] MJ2 = {K1, K2, ..., Km};

[0089] Among them, Jn represents the nth peak screened out by M1, Km represents the mth peak screened out by M2, n represents the maximum peak number screened out by M1, and m represents the maximum peak number screened out by M2. Each element in the set represents the size of the peak and its corresponding timestamp.

[0090] The time difference between MJ1 and MJ2 is used for positioning to determine the approximate direction of the sound source in the earphone.

[0091] Judgment of unilateral sound source: When m=n, according to the number of each element in MJ1 and MJ2, the elements with the same number in MJ1 and MJ2 are matched one-to-one, and the timestamp of each element in MJ1 and MJ2 is obtained; if the timestamp of each element in MJ1 is earlier than the timestamp of the corresponding element in MJ2, the sound source is determined to be on the side of sound sensor M1; if the timestamp of each element in MJ2 is earlier than the timestamp of the corresponding element in MJ1, the sound source is determined to be on the side of sound sensor M2.

[0092] When the judgment of a unilateral sound source is not satisfied, it is judged as a non-unilateral sound source.

[0093] If the sound source is determined to be unilateral, the attenuation of the sound intensity at the sound sensors on both sides is evaluated.

[0094]

[0095] Where i represents the index of the peak number; ρ represents the attenuation rate of the sound intensity from the sensor close to the sound source to the sensor far away from the sound source; I i大 and I i小 Represents the sound intensity of the two sensors at the i-th peak, and I i大 >I i小 .

[0096] What needs to be said is that sound data is collected using sensors on both sides of the headset (M1 and M2), with the aim of accurately determining the direction of the sound source by comparing the differences in sound data on both sides. Basic data is used to perform a more in-depth analysis of the direction of the sound source. Advanced analysis involves extracting peaks from the time-frequency domain images and using these peaks to compare the timing and intensity differences of the sounds sensed by the left and right headsets. By identifying and comparing the sound peaks of the left and right headsets (MJ1 and MJ2), the differences in the timestamps of these peaks can be used to infer the approximate direction of the sound source. When the peak numbers on both sides are equal, the timestamps are compared to determine which headset the sound source is closer to. This is very useful for determining whether the sound source is on the left or right side of the user.

[0097] S2: Analyze the data information and generate feature information.

[0098] The data information is analyzed and feature information is generated, including: if it is determined to be a unilateral sound source, the time-frequency domain data of the sound sensor on the side close to the sound source are analyzed; if it is determined to be a non-unilateral sound source, the time-frequency domain data of the sound sensors on both sides are analyzed respectively.

[0099] The analysis process of time-frequency domain data is completed through neural networks:

[0100] Input layer: Input time-frequency domain image data, which is converted into a format acceptable to the model after preprocessing, denoted as S(t,f).

[0101] TCN layer output:

[0102] TCN output =ReLU(W tcn *S(t,f)+b tcn )

[0103] Among them, W tcn represents the convolution kernel weight in the temporal convolutional network layer, b tcn Represents the bias term in the temporal convolutional network layer.

[0104] LSTM layer output:

[0105] LSTM output =LSTM(TCN output )

[0106] Among them, TCN output Represents the output features at time t after processing by the temporal convolutional network.

[0107] Attention Weighting:

[0108]

[0109] e t represents the attention score at time t:

[0110] e t =v Z tanh(W a LSTM output (t)+b a )

[0111] Where Z represents the transpose; v represents the weight vector of the attention score; W a represents the weight of the attention layer; b a Represents the bias term; τ represents the τth moment predicted from the current moment; T represents the length of the predicted time.

[0112] Weighted summation produces the final feature:

[0113]

[0114] Output prediction:

[0115] [PL, PP] = W0*Attention output +b0

[0116] Among them, W0 represents the output weight matrix, which is used to linearly convert the feature vector output by the attention mechanism into the final prediction target, that is, the predicted loudness and pitch; b0 represents the bias vector of the output model; PL represents the predicted loudness; PP represents the predicted pitch.

[0117] PL and PP are used as feature information of a predicted time length T, and the feature information is used to trigger the control of the interchangeable earplugs of the sleeping headset.

[0118] It's important to note that analyzing time-frequency domain data can capture subtle changes in sound signals. A deep learning framework combining a temporal convolutional network (TCN) and a long short-term memory network (LSTM) with an attention mechanism is used to process and analyze sound data, enabling both in-depth and forward-looking data processing. This approach combines the efficient feature extraction of TCN with the temporal memory capabilities of LSTM, and optimizes feature utilization through the attention mechanism to form predictions of future sound states. The TCN layer processes data using its unique convolution kernel weights and bias terms, effectively capturing dependencies and features in the time series. The LSTM layer processes features after the TCN layer and maintains the flow of important information for long-term sequence prediction. The attention mechanism strengthens the model's focus on specific moments or features, thereby improving prediction accuracy and relevance. All features are integrated through a weighted summation, and the output weight matrix and bias vector are used to produce the final prediction results: the loudness rating (PL) and pitch rating (PP) predicted for the next time period T. Through deep learning of time-frequency domain data through neural networks, more accurate sound analysis and prediction can be achieved than traditional methods, which is of great significance for improving sleep quality and user comfort.

[0119] S3: Using the characteristic information as a trigger signal to adjust the sleep headphones.

[0120] Get the preferred loudness P set by the user user , introduces an adaptive adjustment factor based on predicted loudness and pitch, and a dynamic loudness range compressor to optimize the volume output.

[0121] Volume adjustment factor:

[0122]

[0123] Among them, α, β, and γ represent adjustment parameters, which are used to balance the sensitivity of loudness adjustment and the periodic effect of pitch; P max Indicates the maximum value of pitch, used to normalize pitch values. α is typically set to positive values, such as from 0.1 to 10. Lower values ​​make volume adjustments less sensitive to changes in loudness, while higher values ​​increase this sensitivity. β can be positive or negative, depending on whether you want loudness increases to increase or decrease the effect of volume adjustments. It might range from -2 to 2, where negative values ​​indicate greater suppression of volume at higher loudness levels. γ is typically set to positive values, perhaps from 0.1 to 10, and adjusts the sensitivity of the pitch to periodic effects.

[0124] The specific method for obtaining α, β, and γ is to initially set the parameters and then adjust them based on user listening feedback until the optimal listening experience is achieved. A large amount of user listening feedback and ambient sound data is collected, and machine learning methods (such as regression analysis and gradient descent) are used to optimize the parameter values ​​to minimize prediction error or maximize user satisfaction. An acoustic model is used in a simulated environment to predict the volume output effect under different parameters and select the optimal parameter combination.

[0125] Dynamic loudness range compression:

[0126] V out =V base (1+V factor )·log(1+exp(V factor ))

[0127] Among them, V base Indicates the reference volume.

[0128] If it is judged to be a unilateral sound source, V out To control the target, the volume of the earphone close to the sound source is adjusted; and the volume of the earphone far from the sound source is adjusted synchronously according to the attenuation rate. The adjusted volume is expressed as: ρ·V out If it is determined to be a non-unilateral sound source, the V calculated for the sound on both sides according to their respective feature information is out When adjusting the volume, the headphone volume is adjusted at a preset speed.

[0129] It should be noted that in the case of a unilateral sound source, since only the calculation is performed on the earphones near the sound source, the amount of calculation is greatly reduced and the calculation time is shortened. By intelligently adjusting the volume of the earphones on both sides, one ear can be prevented from being subjected to excessive sound pressure, thereby protecting the user's hearing and reducing auditory fatigue that may be caused by long-term exposure to high-volume environments. In a dynamic environment, the position and nature of the sound source may change frequently. It can automatically identify changes in the sound source and quickly adjust the volume without manual settings by the user, which improves the convenience of the user interface and the adaptability of the system. The preset adjustment speed ensures that the volume changes are not too abrupt or too slow, thereby quickly adapting to changes in the environment while ensuring the naturalness and comfort of the volume adjustment.

[0130] S4: Use the force sensor of the interchangeable earplug to sense the force state of the ear canal on the earplug.

[0131] A miniature force sensor embedded in the interchangeable earbud measures the radial force exerted on the earbud by the ear canal. If the sound source is determined to be unilateral, the radial force applied to the earbud closer to the source is controlled based on the PP and PL values. The radial force applied to the earbud farther from the source is also controlled simultaneously based on the attenuation rate. If the sound source is not determined to be unilateral, the radial force applied to both earbuds is controlled separately.

[0132] S5: Adjusting the pressure of the interchangeable earplug according to the force state and the adjustment result of the sleep earphone.

[0133] Under the predicted environment of PP and PL, the pressure exerted by the interchangeable earplug on the inside of the ear canal is controlled according to the radial force exerted by the inside of the ear canal on the earplug and the volume of the sleeping earphone.

[0134] Stress model;

[0135]

[0136] Where P represents the pressure applied by the earplug in the ear canal; It represents the second derivative of pressure with respect to time and is the acceleration of pressure change; represents the spatial derivative term, c is the propagation speed of the sound wave in the ear canal medium; θP represents the nonlinear damping term of the ear plug material, and θ is the damping coefficient; It represents the intensity coefficient of the sound wave source term, which is related to the energy of the sound emitted by the headphones; V represents the headphone volume; ε represents the nonlinear coefficient that adjusts the loudness and volume; PP0 represents the resonant frequency of the ear canal, and δ represents the attenuation coefficient that describes when the frequency deviates from the resonant frequency.

[0137] Through human-computer interaction, the system is instructed to change the media. When the user replaces the earplug media, the damping coefficient comparison table of the media is used to update θ in the nonlinear damping term. By solving the pressure model, the radial pressure P exerted by the earplug on the ear canal is calculated, and control instructions are issued to control the pressure.

[0138] It is worth mentioning that the coefficients in the pressure model δ and ε are obtained by experimental fitting. The steps to determine the output power density of the headphones include: measuring the output power density P out : Measure the sound pressure level of headphones at maximum volume at a specific distance using a standard sound level meter. Estimating the coupling efficiency η between the headphone and the ear canal: Experimentally determine the sound pressure level transfer efficiency of different earplugs under the same sound source conditions.

[0139]

[0140] Adjust the nonlinear coefficient ε that affects loudness and volume. Through user testing, collect user comfort feedback under different volume and loudness combinations. The fitting parameters are obtained by fitting.

[0141] The steps for determining δ include: determining the ear canal resonant frequency (PP0): using acoustic testing equipment to measure the natural resonant frequency of the ear canal when it is unobstructed. Measuring the attenuation rate when the resonant frequency deviates from the resonant frequency: measuring the sound pressure level at different frequency points near the resonant frequency and recording the attenuation when the resonant frequency deviates from PP0. Fitting the attenuation model: using the measured data, using an appropriate mathematical model (such as an exponential decay model or a Gaussian model) to describe the relationship between frequency and attenuation.

[0142]

[0143] Here, the response ratio refers to the ratio of the sound pressure levels measured at the resonant frequency PP0 and the other frequencies PP.

[0144] Furthermore, the sound sensor of the sleep headset continuously detects external sound information on both sides of the left and right ears. Based on the continuous detection information from both sides of the left and right ears, the volume of the interchangeable earplugs of the sleep headset and the pressure on the ear canal are controlled in real time.

[0145] If the external sound information continuously detected by M1 and the external sound information continuously detected by M2 change simultaneously, and the sound characteristics of M1 during continuous detection are the same as the sound characteristics of M2 before the change, and the sound characteristics of M2 during continuous detection are the same as the sound characteristics of M1 before the change, then it is determined that the left and right interchangeable earbuds of the sleep headset have been swapped. When the left and right interchangeable earbuds of the sleep headset have been swapped, the control data of M1 and M2 are replaced so that the left and right ears can maintain the original volume and pressure. After the replacement is completed, continuous detection and real-time control of M1 and M2 continue.

[0146] It should be noted that by monitoring the intensity and frequency distribution of external sounds in real time, and adjusting the volume of the headphones and the pressure of the earplugs according to these data, the present invention can dynamically adapt to the user's sleeping environment. This intelligent adjustment can significantly reduce the interference of ambient noise on the user's sleep, thereby improving sleep quality. The integrated force sensor monitors the force state of the ear canal on the earplug, allowing the system to adjust the pressure according to real-time data to ensure that the earplug is neither too tight nor too loose, avoiding discomfort or health problems caused by long-term wear. This is an important guarantee of safety and comfort for users. Through the human-computer interaction interface, users can set headphone parameters such as volume, loudness, and pressure according to their personal preferences. In addition, the system can automatically adjust the parameters after the earplug medium is replaced to adapt to the physical properties of the new medium, ensuring that the device always operates in the best condition.

[0147] When it detects that the earbuds are worn incorrectly (such as the left and right earbuds are swapped), the system can automatically adjust the control parameters to ensure that the volume and pressure are set correctly, avoiding affecting the user experience or sleep quality due to earbud replacement.

[0148] On the other hand, this embodiment also provides an adjustment system for interchangeable earplugs of sleep headphones based on external sounds, which includes:

[0149] The acquisition unit obtains data information of external sounds and uses the force sensor of the interchangeable earplug to sense the force state of the ear canal on the earplug.

[0150] The analyzing unit analyzes the data information and generates feature information.

[0151] The control unit uses the characteristic information as a trigger signal to adjust the sleep earphones; and adjusts the radial pressure of the interchangeable earplugs according to the force state and the adjustment result of the sleep earphones.

[0152] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0153] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0154] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0155] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0156] Example 2

[0157] The following is an embodiment of the present invention, which provides a method for adjusting the interchangeable earplugs of sleep headphones based on external sounds. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0158] Experiment 1:

[0159] The experimental subjects included two groups of headphones: one group used the present invention (hereinafter referred to as the "experimental group"), and the other group used common non-smart sleep headphones on the market (hereinafter referred to as the "control group"). Fifty participants in each group used the headphones for one week in the same noise environment.

[0160] The experimental environment simulated a nighttime urban noise environment, including traffic, human voices, and industrial noise, with an average sound intensity of 60dB. All participants slept in this environment and evaluated the headphones during use.

[0161] The experimental earphones are equipped with high-precision sound sensors that capture real-time data about external sounds, including sound intensity, frequency distribution, and the duration and time of events. The earphones process this data using a built-in neural network, identifying noise patterns and automatically adjusting the volume and pressure of the earbuds to maximize noise suppression while maintaining ear canal comfort.

[0162] The control group headphones used traditional passive noise reduction technology, without real-time processing of sound data and dynamic adjustment of earplug pressure.

[0163] Detailed experimental operation:

[0164] Before the experiment began, all participants received training on how to use the earphones correctly to ensure data validity. The experimental group used a human-computer interface to set their personal preferences, including the type of noise they most wanted to suppress and their preferred basic pressure for the earplugs.

[0165] During the experiment, the experimental group's headphones adjusted the volume and pressure of the earbuds based on real-time analysis of external sound data. Force sensors in the earbuds continuously monitored the pressure inside the ear canal to ensure that pressure adjustments did not cause discomfort to the user.

[0166] The headphone system also records all automatic adjustment events and the user's manual adjustment behavior to evaluate the system's intelligent response efficiency and user satisfaction. The recorded results are shown in Table 1 (all statistical results are average values).

[0167] Table 1 Test data table

[0168]

[0169] As can be seen from Table 1, the smart earphones in the experimental group are significantly better than those in the control group in terms of noise suppression, with an average noise suppression of 15dB higher. This shows that the sound analysis and dynamic adjustment technology of the present invention can effectively identify and suppress external noise, providing a quieter sleeping environment. In terms of user comfort, the score of the experimental group is also significantly higher than that of the control group, with an average score of 3 points higher. This shows that the dynamic pressure adjustment function of the earplug can adjust the pressure according to the user's ear canal condition and noise environment, increasing the comfort and applicability of the earphones. The data of the number of automatic adjustment responses shows that the experimental group earphones automatically adjusted 150 times in one week of use, while the control group did not have an automatic adjustment function. This data reflects the dynamic response capability of the smart earphones of the present invention, which can adapt to environmental changes in real time without user intervention. The user satisfaction score further verifies the advantages of the present invention. The average satisfaction score of the experimental group is 9, which is much higher than the 6 points of the control group. This shows that users are very satisfied with the overall experience of the smart earphones, especially the improvement in its noise suppression effect and comfort.

[0170] Experiment 2:

[0171] The experimental subjects were divided into two groups: one group used the smart sleep headphones of the present invention (the experimental group) and the other group used conventional sleep headphones (the control group). The experiment involved 100 adult volunteers who used the headphones during their sleep for five consecutive nights. The volunteers were randomly assigned to the two groups.

[0172] Three different environmental noise scenarios were simulated in the experiment:

[0173] Urban traffic noise (average sound level 75dB).

[0174] Home environment noise (such as TV sound, average sound level 60dB).

[0175] Static environment (basically no noise, sound level about 30dB).

[0176] All participants wore the appropriate headphones during the experiment and recorded their sleep quality, comfort, and next-day sleep state each night. Sleep tracking devices collected sleep data, including time to sleep onset, duration of deep sleep, and number of awakenings during the night. The experimental group's device also automatically recorded the number and type of adjustments made each night. The results are shown in Table 2 (all statistically averaged).

[0177] Table 2 Experimental statistics

[0178]

[0179] The data in Table 2 shows that the experimental group's smart sleep headphones performed better than the control group in both urban traffic noise, home ambient noise, and static environments. The experimental group experienced significantly shorter sleep delays, longer periods of deep sleep, and fewer nighttime awakenings in the urban traffic noise environment. This demonstrates that the intelligent sleep control technology of the present invention can effectively suppress noise interference and improve sleep quality.

[0180] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for adjusting interchangeable earplugs of a sleeping headset based on external sounds, characterized in that: include: Obtain data information of external sounds; Analyzing the data information and generating feature information; Using the characteristic information as a trigger signal to adjust the sleep headphones; The force sensor of the interchangeable earplug is used to sense the force applied by the ear canal to the earplug. adjusting the pressure of the interchangeable earplugs according to the force state and the adjustment result of the sleep earphones; The data information obtained from the external sound includes basic data and advanced data; The basic data includes sound intensity data, frequency distribution data, duration and occurrence time of sound events; The advanced data includes determining the direction of the sound source based on the basic data obtained from the left and right headphones respectively. The specific steps are as follows: Assume that there are sound sensors on both sides of the earphone M1 and M2 respectively, and obtain the time-frequency domain data of M1 and M2; In the time-frequency domain image of each sound sensor, peak recognition is performed on each time-frequency domain image segment, and the average value of the peak amplitude in the current time-frequency domain image segment is calculated. Peaks with amplitudes exceeding the average value are screened out and collected as MJ1 and MJ2 respectively; MJ1={J1,J2,...,Jn}; MJ2 = {K1, K2, ..., Km}; Where Jn represents the nth peak filtered out by M1, Km represents the mth peak filtered out by M2, n represents the maximum peak number filtered out by M1, and m represents the maximum peak number filtered out by M2. Each element in the set represents the size of the peak and its corresponding timestamp. Use the time difference between MJ1 and MJ2 to determine the approximate location of the sound source in the earphones; Judgment of unilateral sound source: When m=n, according to the number of each element in MJ1 and MJ2, the elements with the same number in MJ1 and MJ2 are matched one-to-one, and the timestamp of each element in MJ1 and MJ2 is obtained; if the timestamp of each element in MJ1 is earlier than the timestamp of the corresponding element in MJ2, the sound source is determined to be on the side of sound sensor M1; if the timestamp of each element in MJ2 is earlier than the timestamp of the corresponding element in MJ1, the sound source is determined to be on the side of sound sensor M2; When the judgment of a unilateral sound source is not met, it is judged as a non-unilateral sound source; If the sound source is determined to be unilateral, the attenuation of the sound intensity at the sound sensors on both sides is evaluated; Where i represents the index of the peak number; Indicates the attenuation rate of sound intensity from the sensor close to the sound source to the sensor far away from the sound source; I i大 and I i小 Represents the sound intensity of the two sensors at the i-th peak, and I i大 >I i小 ; Adjusting the pressure of the interchangeable earplug includes controlling the pressure of the interchangeable earplug on the inside of the ear canal according to the radial force of the inside of the ear canal on the earplug and the volume of the sleep earphones under the predicted environment of the pitch and loudness; Stress model; Where P represents the pressure applied by the earplug in the ear canal; It represents the second derivative of pressure with respect to time and is the acceleration of pressure change; represents the spatial derivative term, c is the propagation speed of the sound wave in the ear canal medium; θP represents the nonlinear damping term of the ear plug material, and θ is the damping coefficient; represents the intensity coefficient of the sound wave source term, which is related to the energy of the sound emitted by the headphones; V represents the headphone volume; ε represents the nonlinear coefficient that adjusts the loudness and volume; PP0 represents the resonant frequency of the ear canal, and δ represents the attenuation coefficient that describes when the frequency deviates from the resonant frequency; PL represents the predicted loudness; PP represents the predicted pitch; Through human-computer interaction, the system is instructed to change the medium. When the user changes the earplug medium, the damping coefficient comparison table of the medium is used to update θ in the nonlinear damping term. By solving the pressure model, the radial pressure P exerted by the earplug on the ear canal is obtained, and a control instruction is issued to control the pressure.

2. The method for adjusting interchangeable earplugs for sleep headphones based on external sound according to claim 1, characterized in that: Analyzing the data information and generating characteristic information includes analyzing the time-frequency domain data of the sound sensor on the side close to the sound source if it is determined to be a unilateral sound source; and analyzing the time-frequency domain data of the sound sensors on both sides if it is determined to be a non-unilateral sound source; The analysis process of time-frequency domain data is completed through neural networks: Input layer: Input time-frequency domain image data, which is converted into a format acceptable to the model after preprocessing, denoted as S(t,f); TCN layer output: TCN output =ReLU(W tcn *S(t,f)+b tcn ) Among them, W tcn represents the convolution kernel weight in the temporal convolutional network layer, b tcn Represents the bias term in the temporal convolutional network layer; LSTM layer output: LSTM output =LSTM(TCN output ) Among them, TCN output Represents the output features at time t after processing by the temporal convolutional network; Attention Weighting: e t represents the attention score at time t: e t =v Z tanh(W a LSTM output (t)+b a ) Where Z represents the transpose; v represents the weight vector of the attention score; W a represents the weight of the attention layer; b a Represents the bias term; τ represents the τth moment predicted from the current moment; T represents the length of the predicted time; Weighted summation produces the final feature: Output prediction: [PL,PP]=W0*Attention output +b0 Where W0 represents the output weight matrix, which is used to linearly transform the feature vector output by the attention mechanism into the final prediction target, i.e., the predicted loudness and pitch; b0 represents the bias vector of the output model; PL and PP are used as feature information of a predicted time length T, and the feature information is used to trigger the control of the interchangeable earplugs of the sleeping headset.

3. The method for adjusting the interchangeable earplugs of a sleep headset based on external sound according to claim 2, characterized in that: The adjustment of the sleeping earphones includes obtaining the preferred loudness P set by the user. user , introduces an adaptive adjustment factor based on predicted loudness and pitch, and a dynamic loudness range compressor to optimize volume output; Volume adjustment factor: Among them, α, β, and γ represent adjustment parameters, which are used to balance the sensitivity of loudness adjustment and the periodic effect of pitch; P max Indicates the maximum value of the pitch, used to normalize the pitch value; Dynamic loudness range compression: V out =V base ·(1+V factor )·log(1+exp(V factor )) Among them, V base Indicates the base volume; If it is judged to be a unilateral sound source, V out To control the target, the volume of the earphone close to the sound source is adjusted; and the volume of the earphone far from the sound source is adjusted synchronously according to the attenuation rate. The adjusted volume is expressed as: ρ·V out ; If it is determined to be a non-unilateral sound source, the V calculated for the sound on both sides according to their respective feature information is out Perform volume control; When adjusting the volume, the headphone volume is adjusted at a preset adjustment speed.

4. The method for adjusting the interchangeable earplugs of a sleep headset based on external sound according to claim 3, characterized in that: The method of using the force sensor of the interchangeable earplug to sense the force state of the ear canal on the earplug includes embedding a miniature force sensor in the interchangeable earplug to measure the radial force of the ear canal on the earplug; If the sound source is determined to be unilateral, the radial force applied to the earplug on the side close to the sound source is controlled based on PP and PL; and the radial force applied to the earplug on the side away from the sound source is simultaneously controlled according to the attenuation rate; If it is determined that the sound source is not unilateral, the radial force applied to the earplugs on both sides is controlled separately.

5. The method for adjusting the interchangeable earplugs of a sleep headset based on external sound according to claim 4, characterized in that: Adjusting the pressure of the interchangeable earplugs further includes continuously detecting external sound information on both sides of the left and right ears through a sound sensor of the sleep earphones; Based on the continuous detection information of the left and right ears, the volume and pressure of the interchangeable earbuds of the sleep headphones are controlled in real time; If the external sound information continuously detected by M1 and the external sound information continuously detected by M2 change at the same time, and the sound characteristics of M1 during the continuous detection are the same as the sound characteristics of M2 before the change, and the sound characteristics of M2 during the continuous detection are the same as the sound characteristics of M1 before the change, then it is determined that the left and right earplugs of the interchangeable sleep headphones have been swapped; When the interchangeable earbuds of the sleep headphones are swapped between left and right, the control data of M1 and M2 are replaced so that the left and right ear sides can have the original volume and pressure; after the replacement is completed, continuous detection and real-time control of M1 and M2 continue.

6. A system for adjusting interchangeable earplugs for sleep headphones based on external sound, using the method according to any one of claims 1 to 5, characterized in that: The acquisition unit acquires data information about external sounds and uses the force sensor of the interchangeable earplug to sense the force applied by the ear canal to the earplug; an analysis unit, for analyzing the data information and generating feature information; A control unit, using the characteristic information as a trigger signal to adjust the sleep headphones; According to the stress state and the adjustment result of the sleep earphone, the radial pressure of the interchangeable earplug is adjusted.

Images