Le-audio bluetooth earphone motion state adaptive playing control method, system and device

Abstract

The application discloses a le-audio Bluetooth earphone motion state adaptive playing control method, system and device, relates to the technical field of voice processing, and is applied to a Bluetooth earphone.The method comprises the following steps: S100, dividing an audio signal into a plurality of time windows, and calculating a fitting deviation value in each time window; the size of the time window is set according to a sampling rate of the audio signal and expected analysis accuracy; in each time window, a fitting deviation value at each moment is calculated according to an actual amplitude of the audio signal and a fitting value on a trend term curve; S200, counting the number and distribution of different types of fitting deviation values in each time window to obtain a type density histogram of the fitting deviation values; and S300, fusing the probability density distribution of adjacent time windows by using a weighted average method; and the technical effects of dynamic adjustment, improved denoising accuracy and adaptive capacity are realized.

Description

Technical Field

[0001] This invention relates to the field of voice processing technology, and in particular to a method, system, and device for adaptive playback control of le-audio Bluetooth headphones in motion state. Background Technology

[0002] While Bluetooth headsets are convenient to carry and use, they also have drawbacks. No matter how advanced Bluetooth technology is, its transmission performance is always weaker than wired transmission. During transmission, audio data loss, corruption, and noise frequently occur, leading to audio distortion. Furthermore, the phone's decoding chip also generates slight background noise during audio decoding. In complex environments, such as those with strong electromagnetic interference, Bluetooth headsets may even produce a buzzing noise. While the influence of the external environment cannot be completely avoided, audio distortion can be reduced by optimizing the audio transmission method. Since the signal value of an audio signal can be precisely determined at any given moment, current wired and wireless audio transmissions utilize sampling encoding. Therefore, methods to improve audio quality generally focus on scientifically and rationally adjusting the distribution density of sampling points during encoding and transmission, ensuring signal fidelity while controlling transmission time to reduce noise interference.

[0003] A Chinese invention patent with publication number CN115798502B provides an audio noise reduction method for Bluetooth headsets. This method obtains a trend curve by fitting the amplitude of the acquired audio signal, divides the audio signal into segments based on the trend curve, obtains the fitting deviation value based on the amplitude differences within the audio segments, determines the number of sampling points for each audio segment, and then samples and denoises the audio based on the number of sampling points. However, this prior art ignores the dynamic changes in the audio signal, specifically the distribution and variation of different types of fitting deviation values ​​over time, leading to reduced noise reduction accuracy. Summary of the Invention

[0004] This application provides a method, system, and device for adaptive playback control of le-audio Bluetooth headphones in motion state, which solves the problem in the prior art of ignoring dynamic changes in audio signals, resulting in reduced noise reduction accuracy, and achieves the technical effect of dynamic adjustment, improving noise reduction accuracy and adaptive capability.

[0005] This application provides a motion-state adaptive playback control method for le-audio Bluetooth headphones, applied to Bluetooth headphones, the method comprising:

[0006] S100: Divide the audio signal into multiple time windows and calculate the fitting deviation value within each time window; the size of the time window is set according to the sampling rate of the audio signal and the expected analysis accuracy; within each time window, calculate the fitting deviation value at each moment based on the actual amplitude of the audio signal and the fitted value on the trend term curve.

[0007] S200: Statistically analyze the number and distribution of different types of fitting deviation values ​​within each time window to obtain a type density histogram of fitting deviation values;

[0008] S300: The probability density distributions of adjacent time windows are fused using a weighted average method;

[0009] S400: Arrange the fitting deviation values ​​in each time window according to the normal distribution to obtain the set of deviation values ​​and the serial number corresponding to each fitting deviation value. Use the ratio of the deviation range value to the sequence difference value as the first sampling density of each time window.

[0010] S500: Statistically count the types and quantities of all fitting deviation values ​​in the audio signal, calculate the type influence value, and obtain the second sampling density for each time window; normalize the type influence value of each type of fitting deviation value in the time window, and the second sampling density is the sum of the normalized type influence values.

[0011] Furthermore, the method also includes:

[0012] S600: Use the average of the first and second sampling densities of each time window as the sampling density coefficient; calculate the number of sampling points for each time window based on the proportion of each time window in the audio signal and the sampling density coefficient;

[0013] S700: Samples the audio signal according to the determined number of sampling points, and performs noise reduction processing on the sampled audio signal.

[0014] Furthermore, the method also includes: adaptively adjusting the size of the time window according to the dynamic characteristics of the audio signal, initializing the size of the time window, and for each new time window position, calculating the energy of the audio segment within the time window and dynamically adjusting its size; for high-energy audio signals, decreasing the time window; and for low-energy signals, increasing the time window.

[0015] Furthermore, the method also includes: setting an overlap ratio so that there is a partial overlap area between adjacent time windows; calculating the size of the overlap area between adjacent time windows; calculating the starting point of the subsequent time window based on the starting point of the previous window and the size of the overlap area; and adjusting the size of the last time window to the length of the remaining audio if the total length of the audio signal is not an integer multiple of the time window size.

[0016] Furthermore, the method also includes: adaptively setting a threshold for the type density distribution of fitting deviation values ​​based on the dynamic characteristics of the audio signal; when calculating the type density distribution of fitting deviation values, it is necessary to analyze the changing trend of fitting deviation values ​​within a time window and calculate the slope of the deviation sequence within the time window to obtain the changing trend; according to the set threshold and the changing trend of the deviation, the fitting deviation values ​​are divided into different categories, and the density distribution of each type of deviation value is statistically analyzed.

[0017] Furthermore, the method also includes: acquiring motion state information of the Bluetooth headset, preprocessing and extracting features from the acquired motion state information, determining the motion state of the Bluetooth headset based on the extracted motion features; preprocessing the audio signal, extracting features from the audio signal using audio analysis technology, matching the extracted audio features with features in a motion noise feature library, and identifying the noise type.

[0018] Furthermore, the motion noise feature library needs to collect noise samples generated under various motion states, preprocess and extract features from the noise samples, extract representative noise features, store the extracted noise features in the motion noise feature library, and label the corresponding motion state and noise type.

[0019] Furthermore, the method also includes: setting adjustment rules for the time window size, overlap ratio, and fitting deviation threshold based on the perceived motion state and noise type; dynamically adjusting the time window size and overlap ratio according to the adjustment rules; and dynamically adjusting the threshold of the type density distribution of the fitting deviation value.

[0020] The le-audio Bluetooth headset motion state adaptive playback control system, along with the le-audio Bluetooth headset motion state adaptive playback control method described above, includes: a time window adjustment module for setting the size and adjustment rules of the time window;

[0021] The audio signal segmentation module is used to acquire audio signals and divide them into multiple time windows; it receives the setting rules from the time window adjustment module and performs segmentation according to the rules, while transmitting the segmented time windows to the analysis and statistics module.

[0022] The analysis and statistics module is used to calculate the fit deviation value at each time point and to count the number and distribution of different types of fit deviation values ​​within each time window;

[0023] The noise reduction module is used to calculate the sampling density and sampling density coefficient, and then calculate the number of sampling points to perform noise reduction processing on the sampled audio signal.

[0024] An LE-audio Bluetooth headset motion-adaptive playback control device, coupled with the LE-audio Bluetooth headset motion-adaptive playback control method described above, the device includes a Bluetooth headset with a built-in LE-audio chip and a terminal device, the Bluetooth headset receiving audio signals sent by the terminal device; it also includes a storage device and a processor, the storage device being used to store one or more programs;

[0025] When the processor executes the one or more programs, the processor implements the adaptive playback control method based on the motion state of the le-audio Bluetooth headset described above.

[0026] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0027] By dividing the time window, the dynamic changes of the audio signal can be captured more precisely, and the characteristic distribution of the audio signal in different time periods can be reflected more accurately, thereby achieving a more effective noise reduction effect and improving the noise reduction accuracy and adaptive capability. Attached Figure Description

[0028] Figure 1 This is a flowchart of the le-audio Bluetooth headset motion state adaptive playback control method in an embodiment of the present invention;

[0029] Figure 2 This is the motion state adjustment rule table for the le-audio Bluetooth headset motion state adaptive playback control method in this embodiment of the invention;

[0030] Figure 3 This is a diagram illustrating the architecture of the le-audio Bluetooth headset motion-adaptive playback control system in an embodiment of the present invention. Detailed Implementation

[0031] To facilitate understanding of the present invention, a more complete description of this application will be given below with reference to the accompanying drawings, which illustrate preferred embodiments of the invention. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to enable a more thorough and complete understanding of the disclosure of the present invention.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to limit the invention; the term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0033] Example 1: As Figure 1 As shown, the le-audio Bluetooth headset motion state adaptive playback control method of this application is applied to Bluetooth headsets, and the method includes:

[0034] S100: Divide the audio signal into multiple time windows and calculate the fitting deviation value within each time window; the size of the time window is set according to the sampling rate of the audio signal and the expected analysis accuracy; within each time window, calculate the fitting deviation value at each moment based on the actual amplitude of the audio signal and the fitted value on the trend term curve.

[0035] Specifically, the fit deviation value reflects the degree of difference between the audio signal and the trend term curve, accurately capturing the dynamic changes of different types of fit deviation values ​​over time; the sampling rate represents the number of samples extracted from the continuous signal per second to form the discrete signal, determining the number of sampling points per second in the audio signal; the expected analysis accuracy refers to the desired level of precision or time resolution when analyzing the audio signal; the time window is a fixed time period selected when analyzing the audio signal to capture and analyze specific features or events in the signal. The expected analysis accuracy determines the length of the selected time window. A shorter time window can provide higher time resolution, thus more accurately capturing rapid changes or transient events in the audio signal.

[0036] For example, consider an audio signal with a sampling rate of 44100Hz and an expected analysis precision of 0.01 seconds, meaning the audio signal should be analyzed within 0.01-second time intervals. The time window is determined based on the sampling rate and analysis precision. First, the number of samples per second is calculated; since the sampling rate is 44100Hz, this means 44100 samples per second. Then, the number of samples within each window is determined based on the analysis precision; since the precision is 0.01 seconds, each time window should contain 441 samples. Therefore, each time window is 0.01 seconds in size and contains 441 samples. The size of the time window can be adjusted according to actual needs.

[0037] S200: Statistically analyze the number and distribution of different types of fitting deviation values ​​within each time window to obtain a type density histogram of fitting deviation values;

[0038] Specifically, the probability density distribution within each time window reflects the characteristics of the audio signal during that time period.

[0039] S300: The probability density distributions of adjacent time windows are fused using a weighted average fusion method.

[0040] For example, two adjacent time windows A and B have probability density distributions P and P, respectively. A (x) and P B(x), where x represents the type of fitting deviation value, and the probability density distribution after weighted average fusion can be calculated as:

[0041] P(X) = W A ×P A (x)+W B ×P B (x)

[0042] Among them, W A and W B These are the weighting coefficients for time windows A and B, respectively. The weighting coefficients for time windows can be adjusted based on their importance and degree of change. Within a time window, the greater the energy of the audio segment, the greater its importance and the more drastic its changes.

[0043] By fusing time windows, the probability density distribution between adjacent time windows can transition smoothly, avoiding abrupt changes or discontinuities caused by window division. The fusion process also reduces the impact of noise on the probability density distribution, improving the stability and accuracy of the analysis.

[0044] S400: Arrange the fitting deviation values ​​within each time window according to the normal distribution to obtain the set of deviation values ​​and the corresponding number of each fitting deviation value. Use the ratio of the deviation range value to the sequence difference value as the first sampling density of each time window.

[0045] Specifically, based on the set of deviation values, the difference between the maximum fitted deviation value and the average minimum deviation value is calculated as the deviation range value, and the absolute value of the difference between the maximum sequence number and the left and right minimum sequence numbers is calculated as the sequence difference value.

[0046] S500: Statistically count the types and quantities of all fitting deviation values ​​in the audio signal, calculate the type influence value, and obtain the second sampling density for each time window; normalize the type influence value of each type of fitting deviation value in the time window, and the second sampling density is the sum of the normalized type influence values.

[0047] Specifically, by obtaining the probability density distribution of each type of fitting deviation value, for each time window, any type of fitting deviation value is taken as the reference fitting deviation value, and its type influence value is calculated.

[0048] The type influence value is the result of multiplying the proportion of reference fitting deviation values ​​by their probability density distribution.

[0049] S600: Use the average of the first and second sampling densities of each time window as the sampling density coefficient; calculate the number of sampling points for each time window based on the proportion of each time window in the audio signal and the sampling density coefficient;

[0050] The number of sampling points can be determined by multiplying the proportion of the time window by the sampling density coefficient and combining it with the minimum number of sampling points of the entire audio signal.

[0051] S700: Samples the audio signal according to the determined number of sampling points, and performs noise reduction processing on the sampled audio signal.

[0052] The advantage of using a time window over existing technologies that use audio segments for calculation is that it can capture the dynamic changes of audio signals more precisely. By combining dynamic probability density distribution calculation with time window fusion, it can more accurately reflect the characteristic distribution of audio signals in different time periods, providing a more accurate basis for determining the number of sampling points and noise reduction processing.

[0053] The technical solutions described in the embodiments of this application have at least the following technical effects or advantages:

[0054] This application divides time windows to capture the dynamic changes of audio signals more precisely, and reflects the characteristic distribution of audio signals in different time periods more accurately, thereby achieving a more effective noise reduction effect.

[0055] Example 2: In Example 1, sampling and denoising were performed using a time window. This example is a further improvement on the above example.

[0056] Based on the dynamic characteristics of the audio signal, the size of the time window is adaptively adjusted. The time window size is initialized, and for each new time window position, the energy of the audio segment within the time window is calculated and the size is dynamically adjusted. For high-energy audio signals, the time window is reduced; for low-energy signals, the time window is increased.

[0057] Specifically, the initialization of the time window can be set according to the method of the above embodiment. The energy of the audio segment is obtained by calculating the square of the amplitude of each sampling point, and then accumulating the squared values. The energy value is normalized, and a threshold is set according to the amplitude range of the audio signal. The signal segment with energy greater than the threshold is defined as high energy, and the segment with energy lower than the threshold is defined as low energy. A smaller time window is used in the high-energy part of the audio signal to capture changes more finely, while a larger time window is used in the low-energy part of the signal to reduce the amount of calculation.

[0058] Set the overlap ratio so that there is a partial overlap between adjacent time windows. Calculate the size of the overlap area between adjacent time windows. Calculate the starting point of the subsequent time window based on the starting point of the previous window and the size of the overlap area. If the total length of the audio signal is not an integer multiple of the time window size, adjust the size of the last time window to the length of the remaining audio.

[0059] Specifically, an overlap ratio is set to ensure that even if a feature point is located at the intersection of two time windows, it can be completely contained by at least one time window, capturing continuous changes in the signal and reducing information loss. A certain overlap is set between adjacent time windows, and the overlap ratio can be adjusted as needed. Preferably, an overlap ratio of 50% is recommended.

[0060] For example, if the time window size is 100 milliseconds (assuming a sampling rate of 44.1 kHz, then the corresponding number of sampling points is 4410), and the overlap ratio is 50%, then the size of the overlapping area is: 100 milliseconds * 44.1 kHz * 0.5 = 2205 sampling points.

[0061] Preferably, based on a fixed overlap ratio, the number of sampling points can be calculated to set the window movement step size, i.e., the amount of time the window moves each time. Starting from the beginning of the audio signal, the first time window is applied, and the window size is adjusted according to the aforementioned criteria. The time window is moved along the time axis according to the set movement step size, and the window size is recalculated and adjusted after each movement. This process is repeated until the time window covers the entire audio signal.

[0062] Based on the dynamic characteristics of the audio signal, the threshold for the type density distribution of the fitting deviation value is adaptively set. When calculating the type density distribution of the fitting deviation value, it is necessary to analyze the changing trend of the fitting deviation value within the time window and calculate the slope of the deviation sequence within the time window to obtain the changing trend. Based on the set threshold and the changing trend of the deviation, the fitting deviation value is divided into different categories, and the density distribution of each type of deviation value is statistically analyzed.

[0063] Specifically, a smaller threshold is used in areas of rapid signal change to more finely distinguish different types of fitting deviation values; while a larger threshold is used in areas of relatively stable signal to reduce misjudgments. The slope of the fitting deviation values ​​within the time window is calculated and considered as a factor in the type density distribution. For example, setting the threshold to 0.2, points with an absolute deviation value greater than 0.2 are considered significant deviations. Based on the threshold and the sequence of absolute deviation values, the deviation values ​​can be divided into two categories: significant deviations and non-significant deviations.

[0064] Specifically, calculating the slope of the deviation sequence begins with calculating the difference between the deviation sequences. The difference is the difference between two adjacent deviation values, representing the amount of change in the deviation value. If the time intervals are uniform (i.e., the time interval for each sampling point is the same), then the slope is the average value of the difference sequence; if the time intervals are not uniform, the difference sequence needs to be weighted using the corresponding time intervals, and then the average value is calculated. A positive slope indicates that the deviation value increases with time, a negative slope indicates that the deviation value decreases with time, and a slope close to zero indicates that the deviation value is relatively stable.

[0065] Preferably, the threshold is adaptively set based on the dynamic characteristics of the audio signal. This requires considering the statistical characteristics of the audio signal, such as the mean and standard deviation. An adaptive threshold is set based on the mean and standard deviation. The threshold is set by adding or subtracting one or more standard deviations from the mean. For example, the upper threshold is set to the mean plus 2 standard deviations, and the lower threshold is set to the mean minus 2 standard deviations. When the deviation exceeds these thresholds, it can be considered to represent a significant change or anomaly in the audio signal.

[0066] If the audio signal contains a lot of noise or transient signals, the standard deviation factor needs to be increased to set a stricter threshold. Conversely, if the audio signal is relatively stable, the standard deviation factor can be decreased to set a more lenient threshold.

[0067] In this embodiment, since the characteristics of different audio segments vary in practical applications, time windows of different sizes are required for analysis. The size of the time window is adaptively adjusted according to the dynamic characteristics of the audio signal. A smaller time window is used in parts of the audio signal that change drastically to capture the changes more precisely, while a larger time window is used in parts of the signal that are relatively stable to reduce the amount of computation. A certain overlap ratio is set so that there is some overlap between adjacent time windows, reducing the possibility of information loss.

[0068] The technical solutions described in the embodiments of this application have at least the following technical effects or advantages:

[0069] Based on the dynamic characteristics of audio signals, this application adaptively adjusts the size of the time window and sets overlapping areas to reduce the possibility of information loss, thereby achieving dynamic adjustment of the time window and improving the noise reduction accuracy and adaptability.

[0070] Example 3: In the above examples, the size of the time window is adaptively adjusted according to the dynamic characteristics of the audio signal to achieve the effect of dynamic adjustment. However, in motion, it is necessary to analyze and adjust according to the motion state and noise type.

[0071] The motion state information of the Bluetooth headset is obtained, the obtained motion state information is preprocessed and feature extracted, and the motion state of the Bluetooth headset is determined based on the extracted motion features.

[0072] Specifically, the motion state information of the Bluetooth headset is acquired using an integrated sensor. The Bluetooth headset has a built-in Le-audio Bluetooth chip and transmits data based on Le-audio Bluetooth technology. The acquired motion state information is filtered using median filtering and the data is standardized. Motion features are extracted, including time-domain and frequency-domain features. The time-domain features reflect the changes in motion data over time and can be directly extracted from the raw data. The time-domain features include the mean, variance, and extreme values ​​of acceleration. Fourier transform is used to extract the dominant frequency and frequency distribution range from the frequency-domain features. Based on the motion features, machine learning algorithms are used to classify the motion features and determine the motion state, such as walking, jogging, or running.

[0073] The audio signal is preprocessed, and audio analysis technology is used to extract features from the audio signal. The extracted audio features are then matched with features in the motion noise feature library to identify the noise type.

[0074] The motion noise feature library needs to collect noise samples generated under various motion states, preprocess and extract features from the noise samples, extract representative noise features, store the extracted noise features in the motion noise feature library, and label the corresponding motion state and noise type.

[0075] Based on the perceived motion state and noise type, adjustment rules are set for the time window size, overlap ratio, and fitting deviation threshold. According to the adjustment rules, the time window size and overlap ratio are dynamically adjusted, and the threshold for the type density distribution of fitting deviation values ​​is dynamically adjusted.

[0076] For example, define the motion state: stationary (S), walking (W), jogging (J), running (R); noise type: wind noise (WN), friction noise (FN), other noise (ON); determine the adjustment parameters, time window size (WS), time window overlap ratio (OR), and fit bias threshold (FDT); and formulate an adjustment rule table, such as... Figure 2 As shown, adjustments are made dynamically according to the adjustment rule table.

[0077] Preferably, in high-speed motion or high-noise environments, it is necessary to reduce the time window size and increase the overlap ratio to improve the ability to capture rapidly changing signals or noise. In high-noise environments, it is necessary to increase the threshold to reduce the impact of noise on the fitting results; in scenarios requiring high-precision fitting, it may be necessary to decrease the threshold to improve fitting accuracy.

[0078] The technical solutions described in the embodiments of this application have at least the following technical effects or advantages:

[0079] This application analyzes the motion state and noise type, and dynamically adjusts the time window size, overlap ratio, and fitting deviation threshold, achieving targeted dynamic adjustment under motion conditions.

[0080] Example 4: Based on the above examples, this example provides an adaptive playback control system for the motion state of the le-audio Bluetooth headset, such as... Figure 3 As shown, the system includes: a time window adjustment module, used to set the size and adjustment rules of the time window;

[0081] The audio signal segmentation module is used to acquire audio signals and divide them into multiple time windows; it receives the setting rules from the time window adjustment module and performs segmentation according to the rules, while transmitting the segmented time windows to the analysis and statistics module.

[0082] The analysis and statistics module is used to calculate the fit deviation value at each time point and to count the number and distribution of different types of fit deviation values ​​within each time window;

[0083] The noise reduction module is used to calculate the sampling density and sampling density coefficient, and then calculate the number of sampling points to perform noise reduction processing on the sampled audio signal.

[0084] This embodiment also provides an le-audio Bluetooth headset motion state adaptive playback control device. The device includes a Bluetooth headset with a built-in le-audio chip and a terminal device. The Bluetooth headset receives audio signals sent by the terminal device. It also includes a storage device and a processor. The storage device is used to store one or more programs. When the one or more programs are executed by the processor, the processor implements the le-audio Bluetooth headset motion state adaptive playback control method described above.

[0085] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. For those skilled in the art, the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An adaptive playback control method for motion states in le-audio Bluetooth headphones, applied to Bluetooth headphones, characterized in that, The method includes: S100: Divide the audio signal into multiple time windows and calculate the fitting deviation value within each time window; the size of the time window is set according to the sampling rate of the audio signal and the expected analysis accuracy; within each time window, calculate the fitting deviation value at each moment based on the actual amplitude of the audio signal and the fitted value on the trend term curve. S200: Statistically analyze the number and distribution of different types of fitting deviation values ​​within each time window to obtain a type density histogram of fitting deviation values; S300: The probability density distributions of adjacent time windows are fused using a weighted average method; S400: Arrange the fitting deviation values ​​in each time window according to the normal distribution to obtain the set of deviation values ​​and the serial number corresponding to each fitting deviation value. Use the ratio of the deviation range value to the sequence difference value as the first sampling density of each time window. S500: Statistically count the types and quantities of all fitting deviation values ​​in the audio signal, calculate the type influence value, and obtain the second sampling density for each time window; normalize the type influence value of each type of fitting deviation value in the time window, and the second sampling density is the sum of the normalized type influence values; S600: Use the average of the first and second sampling densities of each time window as the sampling density coefficient; calculate the number of sampling points for each time window based on the proportion of each time window in the audio signal and the sampling density coefficient; S700: Samples the audio signal according to the determined number of sampling points, and performs noise reduction processing on the sampled audio signal.

2. The le-audio Bluetooth headset motion state adaptive playback control method as described in claim 1, characterized in that, The method further includes: adaptively adjusting the size of the time window according to the dynamic characteristics of the audio signal; initializing the size of the time window; and for each new time window position, calculating the energy of the audio segment within the time window and dynamically adjusting its size; decreasing the time window for high-energy audio signals and increasing the time window for low-energy signals.

3. The le-audio Bluetooth headset motion state adaptive playback control method as described in claim 2, characterized in that, The method further includes: setting an overlap ratio so that there is a partial overlap between adjacent time windows; calculating the size of the overlap between adjacent time windows; calculating the starting point of the subsequent time window based on the starting point of the previous window and the size of the overlap; and adjusting the size of the last time window to the length of the remaining audio if the total length of the audio signal is not an integer multiple of the time window size.

4. The le-audio Bluetooth headset motion state adaptive playback control method as described in claim 2, characterized in that, The method further includes: adaptively setting a threshold for the type density distribution of fitting deviation values ​​based on the dynamic characteristics of the audio signal; when calculating the type density distribution of fitting deviation values, it is necessary to analyze the changing trend of fitting deviation values ​​within a time window and calculate the slope of the deviation sequence within the time window to obtain the changing trend; according to the set threshold and the changing trend of the deviation, the fitting deviation values ​​are divided into different categories, and the density distribution of each type of deviation value is statistically analyzed.

5. The le-audio Bluetooth headset motion state adaptive playback control method as described in claim 1, characterized in that, The method further includes: acquiring motion state information of the Bluetooth headset, preprocessing and extracting features from the acquired motion state information, determining the motion state of the Bluetooth headset based on the extracted motion features; preprocessing the audio signal, extracting features from the audio signal using audio analysis technology, matching the extracted audio features with features in a motion noise feature library, and identifying the noise type.

6. The le-audio Bluetooth headset motion state adaptive playback control method as described in claim 5, characterized in that, The motion noise feature library needs to collect noise samples generated under various motion states, preprocess and extract features from the noise samples, extract representative noise features, store the extracted noise features in the motion noise feature library, and label the corresponding motion state and noise type.

7. The le-audio Bluetooth headset motion state adaptive playback control method as described in claim 6, characterized in that, The method further includes: setting adjustment rules for the time window size, overlap ratio, and fitting deviation threshold based on the perceived motion state and noise type; dynamically adjusting the time window size and overlap ratio according to the adjustment rules; and dynamically adjusting the threshold of the type density distribution of the fitting deviation value.

8. A motion-adaptive playback control system for le-audio Bluetooth headphones, coupled with the motion-adaptive playback control method for le-audio Bluetooth headphones as described in any one of claims 1 to 7, characterized in that, The system includes: a time window adjustment module, used to set the size and adjustment rules of the time window; The audio signal segmentation module is used to acquire audio signals and divide them into multiple time windows; it receives the setting rules from the time window adjustment module and performs segmentation according to the rules, while transmitting the segmented time windows to the analysis and statistics module. The analysis and statistics module is used to calculate the fit deviation value at each time point and to count the number and distribution of different types of fit deviation values ​​within each time window; The noise reduction module is used to calculate the sampling density and sampling density coefficient, and then calculate the number of sampling points to perform noise reduction processing on the sampled audio signal.

9. A motion-adaptive playback control device for le-audio Bluetooth headphones, coupled with the motion-adaptive playback control method for le-audio Bluetooth headphones as described in any one of claims 1 to 7, characterized in that, The device includes a Bluetooth headset with a built-in le-audio chip and a terminal device, the Bluetooth headset receiving audio signals sent by the terminal device; it also includes a storage device and a processor, the storage device being used to store one or more programs; When the processor executes the one or more programs, the processor implements the le-audio Bluetooth headset motion state adaptive playback control method according to any one of claims 1 to 7.

Images