Audio adjustment method, apparatus and device based on physiological signals

Abstract

This application discloses an audio adjustment method, apparatus, and device based on physiological signals, applied in the field of audio processing technology. The method involves acquiring the current physiological signals of a target object; extracting multi-dimensional physiological state features based on these signals; determining the current state of the target object according to these features; and dynamically adjusting at least one audio parameter of the audio to be played based on the current state to generate and output audio that matches the current state. This method acquires physiological signals in real time and extracts quantified physiological state features from them to assess the current physiological state, providing a basis for subsequent audio adjustment. Furthermore, by extracting multi-dimensional, deep-level physiological state features, the current state of the target object can be accurately characterized, allowing for dynamic, personalized, and precise adjustment of the audio parameters based on this state.

Description

Technical Field

[0001] This application relates to the field of audio processing technology, and in particular to an audio modulation method, apparatus and device based on physiological signals. Background Technology

[0002] With the increasing popularity of smart mobile devices and the growth of audio content consumption, users' demand for personalized and adaptive audio experiences is growing stronger.

[0003] In related technologies, significant limitations remain in achieving deep, personalized audio adaptation. On one hand, systems often rely on users' historical behavior or explicit settings, lacking the ability to objectively perceive and interpret users' real-time physiological and psychological states, leading to a disconnect between recommendations or adjustments and users' true intrinsic needs. On the other hand, even when certain physiological signals are introduced, their application often remains at a superficial level. Furthermore, existing technologies for adjusting the audio content itself are mostly limited to track switching or overall parameters such as volume and sound effects, resulting in passive, superficial, and limited personalized effects of automated adjustment.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] The purpose of this specification is to provide an audio adjustment method, apparatus, and device based on physiological signals, so that the device can accurately characterize the current state of the target object and achieve dynamic, personalized, and precise adjustment of the audio to be played.

[0006] To address the aforementioned technical problems, this specification provides, in its first aspect, an audio modulation method based on physiological signals, comprising: Obtain the current physiological signals of the target object; Based on the current physiological signal, physiological state features of multiple dimensions are extracted; Based on the physiological state characteristics of the multiple dimensions, the current state of the target object is determined; Based on the current state, at least one audio parameter of the audio to be played is dynamically adjusted to generate and output the audio to be played that matches the current state.

[0007] In some embodiments of this specification, the current physiological signal includes R-wave interval data, and the physiological state characteristics include at least two of the following: time-domain characteristics, frequency-domain characteristics, and nonlinear characteristics; The time-domain features include at least one of the following: statistical features of R-wave interval data, root mean square of the difference between adjacent R-wave intervals, proportion of adjacent R-wave interval differences greater than a preset threshold, and geometric features of the distribution of adjacent R-wave intervals. The frequency domain features include at least one of the following: the frequency band power of the frequency domain signal corresponding to the R-wave inter-interval data in multiple preset frequency bands, the total power of the frequency domain signal, the normalized power of the frequency domain signal, and the ratio of the power of at least two frequency bands. The nonlinear features include at least one of the following: sample entropy of R-wave interval data, self-similarity of R-wave interval data, geometric features of the Poincaré diagram corresponding to R-wave interval data, and complexity of R-wave interval data.

[0008] In some embodiments of this specification, the physiological state features include at least nonlinear features; based on the current physiological signal, multiple dimensions of physiological state features are extracted, including: The R-wave interval data are matched based on a preset sample template to calculate the sample entropy of the R-wave interval data; Based on the R-wave interval data, the detrended fluctuation function corresponding to the R-wave interval data is calculated using the piecewise integration method, and the scaling exponent of the detrended fluctuation function is calculated as the self-similarity of the R-wave interval data. The Poincaré diagram is generated using an ellipse fitting method based on the R-wave interval data. The geometric features of the Poincaré diagram are calculated, and the geometric features include at least one of the following: the major axis, minor axis, and area of ​​the ellipse. Calculate the first complexity of the R-wave interval data at different time scales, the permutation entropy of the R-wave interval data, and the second complexity of the R-wave interval sequence corresponding to the R-wave interval data, and use the first complexity, the permutation entropy, and the second complexity as the complexity of the R-wave interval data.

[0009] In some embodiments of this specification, determining the current state of the target object based on the multiple dimensions of physiological state characteristics includes: The feature weights of physiological state characteristics in each dimension are determined, and the current stress index of the target object is calculated as the current state based on the physiological state characteristics and feature weights of each dimension.

[0010] In some embodiments of this specification, the physiological state characteristics of each dimension include index values ​​corresponding to at least two physiological state indicators; Accordingly, the feature weights of physiological state characteristics in each dimension are determined. Based on the physiological state characteristics and feature weights in each dimension, the current stress index of the target object is calculated as the current state, including: The weights of each physiological state indicator are determined, and the current stress index is calculated as the current state based on the indicator values ​​and weights of each physiological state indicator.

[0011] In some embodiments of this specification, the above method further includes: Obtain the pressure index baseline corresponding to the target object. The pressure index baseline includes multiple preset pressure states and pressure index intervals corresponding to each pressure state. Based on the current pressure index and the pressure index baseline, the current pressure state of the target object is determined as the current state.

[0012] In some embodiments of this specification, the pressure index baseline further includes physiological state characteristic intervals corresponding to each pressure state; the method further includes: Based on the pressure index baseline, determine the initial pressure state corresponding to the current pressure index; When the initial pressure state is determined to be in the boundary region of the pressure state category, a reference pressure state corresponding to the physiological state characteristic is determined based on the pressure index baseline. The current pressure state of the target object is determined as the current state based on the reference pressure state and the initial pressure state.

[0013] In some embodiments of this specification, obtaining the baseline pressure index corresponding to the target object includes: Select a portion of the historical physiological signals of the target object that meet a preset time threshold from the current time as the target physiological signal set; Calculate the historical stress index corresponding to each historical physiological signal in the target physiological signal set, and determine the distribution characteristics of the historical stress index; Based on the aforementioned distribution characteristics, the pressure index range corresponding to each pressure state is determined.

[0014] In some embodiments of this specification, dynamically adjusting at least one audio parameter of the audio to be played includes at least one of the following: Adjust the harmony of the audio to be played; Adjust the volume and / or frequency of each sound element in the audio to be played; Adjust the rhythm parameters of the audio to be played, wherein the rhythm parameters include at least one of the following: beat frequency and beat interval duration.

[0015] In some embodiments of this specification, adjusting the harmonic harmony of the audio to be played includes: Determine the harmonic structure of the audio to be played, wherein the harmonic structure includes a chord sequence composed of multiple chord elements and the chord parameters of each chord element; Based on the mapping relationship between chords and scores, the harmonic weights corresponding to each chord, and the harmonic structure, the harmonic harmony of the audio to be played is determined. Determine the target harmony degree corresponding to the current state; Based on the difference between the stated harmonic harmony and the target harmony, a progressive conversion is used to adjust the harmonic structure of the audio to be played.

[0016] In some embodiments of this specification, the harmonic harmony of the audio to be played is determined based on the mapping relationship between each harmony and the score, the harmonic weights corresponding to each harmony, and the harmonic structure, including: Obtain the functional weight of each chord element in the chord sequence; Based on the chord sequence and the chord parameters of each chord element, calculate the context weight of each chord element; Based on the mapping relationship, the score corresponding to each chord element is determined; The harmonic harmony of the audio to be played is calculated based on the score corresponding to each chord element, the functional weight of each chord element, and the context weight of each chord element.

[0017] In some embodiments of this specification, adjusting the volume and / or frequency of each sound element of the audio to be played includes: The instruments contained in the audio to be played are identified as the first sound elements; the instrument index corresponding to each instrument is obtained, the instrument index being used to characterize the degree of physiological influence of the corresponding instrument on the target object; based on the mapping relationship between stress state and instrument index, and in conjunction with preset volume adjustment rules, the volume of each instrument is adjusted; and / or, The spectrum of the audio to be played is determined as a second sound element. Based on the mapping relationship between the second sound element, pressure state, and frequency band gain value, the frequency band gain value of the audio to be played is adjusted; and / or, The spectral characteristics of each instrument in the audio to be played are determined as the third sound element. Based on the mapping relationship between the third sound element, the pressure state and the frequency band gain value of each instrument, the frequency band gain value of each instrument is adjusted.

[0018] In some embodiments of this specification, adjusting the rhythm parameters of the audio to be played includes: The multiple beat points contained in the audio to be played are determined, and the beat frequency of the audio to be played is determined based on the multiple beat points; Obtain the target beat frequency range corresponding to the current state; adjust the beat frequency based on the target beat frequency range, the beat frequency, and a preset beat frequency adjustment rule; and / or, Calculate the beat interval duration for each pair of beat points; determine the standard deviation and coefficient of variation for multiple beat points based on the beat interval duration; adjust at least one beat interval duration based on the difference between the beat frequency and the target beat frequency range; and / or, Obtain the beat heart rate synchronization rule corresponding to the current state, and adjust the beat frequency based on the beat heart rate synchronization rule.

[0019] The second aspect of this specification provides an audio modulation device based on physiological signals, comprising: The acquisition module is used to acquire the current physiological signals of the target object; The extraction module is used to extract physiological state features in multiple dimensions based on the current physiological signal. The determination module is used to determine the current state of the target object based on the multiple dimensions of physiological state characteristics; An adjustment module is used to dynamically adjust at least one audio parameter of the audio to be played according to the current state, so as to generate and output the audio to be played that matches the current state.

[0020] A third aspect of this specification provides an electronic device, comprising: a memory and a processor, the processor and the memory being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to implement the steps of the method described in the first aspect.

[0021] A fourth aspect of this specification provides a computer-readable storage medium storing computer program instructions that, when executed by a processor, implement the steps of the method described in the first aspect.

[0022] The audio adjustment method, apparatus, and device based on physiological signals provided in the embodiments of this specification acquire the current physiological signals of a target object; extract multiple dimensions of physiological state features based on the current physiological signals; determine the current state of the target object according to the multiple dimensions of physiological state features; and dynamically adjust at least one audio parameter of the audio to be played according to the current state to generate and output the audio to be played that matches the current state. The embodiments of this specification acquire physiological signals in real time and extract quantified physiological state features from the physiological signals to assess the current physiological state, providing a basis for subsequent audio adjustment; and by extracting multiple dimensions and deep-level physiological state features, the current state of the target object can be accurately characterized, and then the audio parameters of the audio to be played can be adjusted based on the current physiological and psychological state, achieving dynamic, personalized, and precise adjustment of the audio to be played, thus improving the user experience. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 The diagram shown is a schematic of an audio modulation method based on physiological signals provided in an embodiment of this specification. Figure 2 The diagram shown is a schematic representation of a nonlinear feature extraction method provided in an embodiment of this specification. Figure 3 The diagram shown is a schematic representation of a method for obtaining a pressure index baseline provided in an embodiment of this specification. Figure 4 The diagram shown is a schematic representation of a method for determining the harmonic structure provided in an embodiment of this specification. Figure 5 The diagram shown is a schematic of an audio modulation device based on physiological signals provided in an embodiment of this specification. Figure 6 The diagram shown is a schematic of an electronic device provided in an embodiment of this specification. Detailed Implementation

[0025] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0026] It should be noted that the information and data related to users involved in the embodiments of this specification are all information and data authorized by the user or fully authorized by the relevant parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with relevant laws, regulations, and standards, and necessary confidentiality measures have been taken. They do not violate public order and good morals, and corresponding operation entry points are provided for users or relevant parties to choose to authorize or refuse.

[0027] It should also be noted that in the embodiments of this specification, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.

[0028] The audio modulation method based on physiological signals provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0029] Figure 1 The diagram illustrates an audio modulation method based on physiological signals provided in an embodiment of this specification. While this specification provides method operation steps or apparatus structures as shown in the following embodiments or figures, the method or apparatus may include more or fewer operation steps or module units through conventional or non-inventive means. In steps or structures where there is no logically necessary causal relationship, the execution order of these steps or the module structure of the apparatus is not limited to the execution order or module structure shown in the embodiments or figures of this specification. When the method or module structure is applied in actual devices, servers, or terminal products, it can be executed sequentially or in parallel according to the method or module structure shown in the embodiments or figures (e.g., in a parallel processor or multi-threaded processing environment, or even in a distributed processing or server cluster implementation environment). Figure 1 As shown, the method may include: S101: Obtain the current physiological signals of the target object.

[0030] It is understood that the target object can be the user, and the current physiological signal can be a real-time biological signal that reflects the state of the user's autonomic nervous system and is collected in a non-invasive manner. Examples include heart rate signals, pulse signals, electroencephalogram (EEG) signals, behavioral and body movement signals, muscle activity signals, and respiratory signals. The current biological signal can be collected by a smart device with biosensing capabilities, and the raw data can be directly used as the current physiological signal, or the raw data can be preprocessed to obtain the current physiological signal. For ease of explanation, this specification uses heart rate signals as an example to describe the above-mentioned audio modulation method based on physiological signals.

[0031] In practice, a smartwatch (such as one integrating a PPG heart rate sensor, a three-axis accelerometer, and a Bluetooth 5.0 module) can be used as the data acquisition terminal, while a mobile device (such as a smartphone or tablet, which can be equipped with a high-performance audio processing chip) can act as the processing and control terminal to execute the aforementioned audio adjustment method. First, after the mobile device is powered on, it can actively scan for smartwatches within its communication range and establish a connection with the target watch via the Bluetooth Low Energy (BLE) protocol. During the connection process, the device model (e.g., confirming that the watch supports a PPG sampling rate ≥100Hz) and functional support (e.g., whether it can output raw RR interval data) are verified. After the connection is established, the mobile device sends a data acquisition command to the watch, which continuously acquires heart rate data at a certain frequency and transmits the heart rate data to the mobile device. Data transmission uses a packet-based mechanism; each transmitted data packet can contain a timestamp, heart rate value, RR interval data (the time interval between two adjacent R waves on an electrocardiogram, in milliseconds), and a signal quality index (range 0-100, with 80 and above considered high-quality data). For example, each data packet is 64 bytes in size. The first 4 bytes are a timestamp, the middle 2 bytes are the heart rate value, the next 8 bytes are the R-wave interval data (or RR interval data), the last 1 byte is the signal quality indicator, and the remaining bytes are checksums. After receiving the data, the mobile device verifies the integrity of the data packet using the CRC32 checksum algorithm. If the verification fails, a retransmission request is sent to the watch, with a maximum of 3 retransmissions. If it still fails, the data is marked as invalid. Simultaneously, the mobile device monitors the Bluetooth connection status in real time, judging connection stability by detecting the Received Signal Strength Indicator (RSSI). When the RSSI is below -80dBm, a reconnection mechanism is automatically initiated, for example, with reconnection intervals of 1s, 3s, 5s, 10s, and 15s, with a maximum of 5 retries. If reconnection fails, a connection interruption alert is sent to the user. Furthermore, the mobile device can maintain a connection status log, recording events such as connection establishment, interruption, and reconnection.

[0032] It is understood that the mobile device is used as an example of the audio adjustment method in the embodiments of this specification. The following embodiments all use the mobile device as the electronic device to apply the method. Other devices may be used in other embodiments, and this specification does not limit them.

[0033] In some embodiments of this specification, obtaining the current physiological signal of the target object includes: obtaining current heart rate data; preprocessing the current heart rate data, the preprocessing including at least one of the following: outlier detection and removal, motion artifact removal, baseline drift correction, and data smoothing; and extracting R-wave interval data as the current physiological signal based on the preprocessed current heart rate data.

[0034] As can be understood, R-wave interval data (also known as RR interval data) refers to the time interval between two adjacent R-wave peaks in an electrocardiogram. It is the core data reflecting heart rate variability (HRV) and can effectively characterize the activity state of the autonomic nervous system.

[0035] In practice, the current heart rate data can be data sent from the watch to the mobile device. The current heart rate data can include photoplethysmography (PPG) data, acquisition timestamps, and signal quality indicators.

[0036] In specific implementation, outlier detection and removal may include: performing statistical analysis on the raw heart rate data based on the 3σ principle, calculating the data mean μ and standard deviation σ, and marking data points exceeding the range of [μ-3σ, μ+3σ] as isolated outliers; for cases where three or more consecutive data points exceed the range, using a sliding window method (with a window length of 5-10 data points) to perform continuous outlier detection, avoiding misjudgments caused by a single outlier; and for marked outliers, using linear interpolation to fill in the gaps based on 2-3 valid data points before and after the outlier.

[0037] In practice, motion artifact removal can include: synchronously acquiring data from the triaxial accelerometer integrated into the smart wearable device; when the effective value of acceleration (RMS) is greater than 0.5g (g is the acceleration due to gravity), the corresponding time period is determined to be a period of intense exercise, and the heart rate data for that time period is marked as unreliable; for the data during periods of intense exercise, an adaptive Kalman filter algorithm is used to dynamically adjust the filter coefficients to suppress artifacts; for periods of mild exercise with an effective value of acceleration between 0.1g and 0.5g, a motion compensation algorithm is used to correct the heart rate signal based on the acceleration data to restore the true physiological signal.

[0038] In practice, baseline drift correction may include: identifying baseline drift components in the heart rate signal through trend analysis, removing slowly changing baseline drift using second- or third-order polynomial fitting methods, and correcting rapidly fluctuating baseline drift using a high-pass filter with a cutoff frequency of 0.05 Hz, while maintaining the short-term variability characteristics of the heart rate signal during the correction process.

[0039] In practice, data smoothing can include: using a Savitzky-Golay filter to smooth the corrected heart rate signal. The filter window size is dynamically adjusted according to the signal quality indicator. When the signal quality indicator is ≥80 (0-100 points), a small window of 5-7 data points is used, and when the signal quality indicator is <80, a large window of 9-11 data points is used. The polynomial order is set to 2nd order to ensure that the signal phase characteristics remain unchanged and the physiological meaning is preserved after smoothing.

[0040] Furthermore, after preprocessing, an adaptive thresholding method can be used to detect the R-wave peaks in the preprocessed heart rate signal. The R-wave peak characteristics are tracked by a sliding threshold. The initial threshold value is set to 50%-60% of the maximum signal value, and the threshold is dynamically adjusted based on the first 5-10 valid R-wave peaks. The difference in the acquisition timestamps corresponding to two adjacent R-wave peaks is calculated to obtain the R-wave interval data. The validity of the extracted R-wave interval data is verified, and data with interval values ​​exceeding the range of 300ms-2000ms are removed. If there are fewer than 5 consecutive valid interval data, re-acquisition or interpolation is triggered.

[0041] In practice, after obtaining the aforementioned R-wave interval data, the R-wave interval data can also be cached based on a data caching mechanism. The data caching mechanism may include: (1) Sliding window management: The mobile device maintains a 5-minute heart rate data sliding window, which supports real-time updates; the window adopts a circular buffer structure to improve memory usage efficiency; when new data arrives, it automatically overwrites the oldest data to keep the window size constant; the mobile device supports multiple windows of different sizes for analysis at different time scales.

[0042] (2) Data integrity check: The mobile device performs integrity verification on each data packet, checking the data format and content; for missing data points, the mobile device records the missing location and reason; when the missing data exceeds the threshold, the mobile device triggers a data quality warning; the mobile device maintains data quality indicators, including integrity, continuity, accuracy, etc.

[0043] (3) Cache optimization strategy: Mobile devices use the LRU (Least Recently Used) algorithm to manage cached data; for frequently accessed data, mobile devices provide a fast access interface; mobile devices support data compression storage to reduce memory usage; cached data is backed up to persistent storage regularly to prevent data loss.

[0044] S102: Based on the current physiological signal, extract physiological state features in multiple dimensions.

[0045] It is understandable that multi-dimensional physiological state features can be a set of features extracted from current physiological signals that comprehensively reflect the autonomic nervous system function of the target object. These features could include two dimensions from time-domain features, frequency-domain features, and nonlinear features, and each dimension could include multiple features. Combining multiple dimensional features can avoid the limitations of a single feature and improve the accuracy of state assessment. In practice, preprocessed R-wave interval data can be used as the basis for extracting physiological state features.

[0046] In some embodiments of this specification, the current physiological signal may include R-wave interval data, and the physiological state characteristics may include at least two of the following: time-domain characteristics, frequency-domain characteristics, and nonlinear characteristics. The time-domain characteristics may include at least one of the following: statistical characteristics of the R-wave interval data, root mean square (RMSSD) of the difference between adjacent R-wave intervals, the proportion of adjacent R-wave interval differences greater than a preset threshold (pNN50), and the geometric characteristics of the distribution of adjacent R-wave intervals. The frequency-domain characteristics may include at least one of the following: the frequency power of the frequency domain signal corresponding to the R-wave interval data in multiple preset frequency bands (e.g., VLF band (0.003-0.04 Hz), LF band (0.04-0.15 Hz), HF band (0.15-0.4 Hz)), the total power of the frequency domain signal, the normalized power of the frequency domain signal, and the ratio of the power of at least two frequency bands (e.g., the LF / HF ratio). The nonlinear features may include at least one of the following: sample entropy of R-wave interval data, self-similarity of R-wave interval data, geometric features of the Poincaré diagram corresponding to R-wave interval data, and complexity of R-wave interval data.

[0047] The statistical characteristics of R-wave interval data can include basic statistics of R-wave intervals, such as mean, standard deviation, coefficient of variation, skewness, kurtosis, etc., and can also include the standard deviation of all RR interval data (SDNN).

[0048] S103: Determine the current state of the target object based on the physiological state characteristics of the multiple dimensions.

[0049] In practical implementation, the physiological state characteristics of the target object across multiple dimensions can be fused and transformed into a comprehensive state label, i.e., the current state, that can directly guide audio adjustment. The current state can characterize the physiological and psychological state of the target object, such as stress levels. The current state can be represented by a quantified numerical value or by a state type; this specification does not impose any restrictions on this. The fusion of multiple physiological state characteristics can be achieved by weighting the values ​​of multiple indicators representing these physiological state characteristics.

[0050] S104: Based on the current state, dynamically adjust at least one audio parameter of the audio to be played in order to generate and output the audio to be played that matches the current state.

[0051] It is understandable that audio parameters can be key indicators that determine the characteristics of audio content. For example, they can include harmony, volume and frequency of sound elements, rhythm parameters, etc. Dynamic adjustment can adjust the above audio parameters in real time according to the current state of the target object, so that the audio to be played can adapt to the physiological and psychological needs of the target object and achieve proactive adjustment.

[0052] In practice, the mobile device can receive the audio stream to be played and analyze its musical characteristics in real time. Based on the determined current state, the corresponding adjustment strategy is retrieved from a preset mapping rule base and applied to the audio stream. The adjustment can target the following three aspects: 1) Harmony: If the current state is under high pressure, the mobile device attempts to identify dissonant chords in the music (such as diminished triads) and gradually replace them with more harmonious chords (such as minor triads) in real time to improve the overall harmony and create a sense of calm.

[0053] 2) Timbre and frequency balance: Identifies the timbre of highly stimulating instruments in the audio (such as distorted electric guitars and sharp cymbals) and automatically attenuates their volume or energy in the high-frequency range; at the same time, it can enhance the volume of soothing instruments (such as acoustic pianos and string backings) or their energy in the mid-low frequency range.

[0054] 3) Rhythm Parameters: The beat frequency (BPM) of the music is detected. If the current state is high-pressure and the music BPM is too fast, the mobile device can use a time-stretching algorithm (such as PSOLA) to gently reduce the playback speed while keeping the pitch constant, slowing the tempo to the target BPM range (such as 60-70 BPM). All adjustments are completed in real time by the audio processing engine, and techniques such as crossfading are used to ensure smooth and imperceptible adjustments. Finally, the adjusted audio stream is sent to speakers or headphones for playback, forming a complete closed loop from physiological signal perception to audio feedback adjustment.

[0055] In the embodiments of this specification, physiological signals are acquired in real time, and quantitative physiological state features are extracted from the physiological signals to assess the current physiological state, providing a basis for subsequent audio adjustment. Furthermore, by extracting multiple dimensions and deep-level physiological state features, the current state of the target object can be accurately characterized. Based on the current physiological and psychological state, the audio parameters of the audio to be played can be adjusted, enabling dynamic, personalized, and precise adjustment of the audio to be played, thereby improving the user experience.

[0056] In some embodiments of this specification, temporal feature extraction may include: (1) RR interval statistical calculation: The mobile device first extracts the RR interval sequence from the heart rate data, that is, the time interval between adjacent R waves; calculates the basic statistics of the RR interval: mean, standard deviation, coefficient of variation, skewness, and kurtosis; the above statistics reflect the basic characteristics and distribution characteristics of heart rate variability; the mobile device normalizes the statistics to eliminate the influence of individual differences.

[0057] (2) SDNN calculation (standard deviation): SDNN is the standard deviation of all RR interval data, reflecting the overall level of heart rate variability; mobile devices use the standard deviation formula to calculate the SDNN value. The larger the SDNN value, the higher the heart rate variability and the better the function of the autonomic nervous system; the mobile device system standardizes the SDNN value according to factors such as age and gender.

[0058] In practice, SDNN values ​​can be obtained through the formula... The calculation shows that RRi can represent the i-th R-wave interval data, RRmean can represent the average value of the R-wave interval data, and N can represent the number of R-wave intervals.

[0059] (3) RMSSD Calculation (Root Mean Square of the Difference Between Adjacent RR Intervals): RMSSD is the root mean square of the squares of the differences between adjacent RR intervals, mainly reflecting parasympathetic activity. The mobile device calculates all adjacent RR interval differences, and then calculates the mean and root mean square of their squares. The higher the RMSSD value, the stronger the parasympathetic activity and the better the heart rate regulation ability. The mobile device performs a logarithmic transformation on the RMSSD value to improve the normality of the data.

[0060] In practice, the RMSSD value can be obtained through the formula. The calculation shows that RRi+1 can represent the (i+1)th R-wave interval data.

[0061] (4) pNN50 calculation (percentage of adjacent RR interval differences > 50ms): pNN50 is the percentage of adjacent RR interval differences exceeding 50 milliseconds, reflecting high-frequency variability. Mobile devices statistically analyze all adjacent RR interval differences and calculate the proportion exceeding 50ms. The higher the pNN50 value, the richer the heart rate variability and the more active the autonomic nervous system regulation. The mobile device system performs an arctangent transformation on the pNN50 value to improve data stability.

[0062] In practice, the pNN50 value can be calculated using the formula pNN50 = Σ(|RRi+1 - RRi| > 50ms) / (N-1) × 100%.

[0063] (5) Geometric index calculation: The mobile device calculates the HRV triangular index, which is the ratio of the height to the width of the RR interval histogram; and calculates TINN (triangular interpolation RR interval histogram baseline width). The above geometric indices provide the geometric characteristics of the heart rate variability distribution, and the mobile device uses interpolation methods to improve the calculation accuracy of the geometric indices.

[0064] In some embodiments of this specification, frequency domain feature extraction may include: (1) Data preprocessing: The mobile device resamples the RR interval sequence to a uniform time series of 4Hz; the cubic spline interpolation method is used for resampling to maintain the smoothness of the signal; the resampled data is detrended to remove long-term trend changes; the Hanning window function is applied to reduce spectral leakage and improve the accuracy of frequency domain analysis.

[0065] (2) FFT Transform and Power Spectrum Calculation: The mobile device uses Fast Fourier Transform (FFT) to convert the time-domain signal into a frequency-domain signal; calculates the power spectral density to obtain the energy distribution of different frequency components; and smooths the power spectrum to reduce the impact of noise. The mobile device system uses the Welch method to improve the stability of power spectrum estimation.

[0066] (3) Frequency band power extraction: VLF band (0.003-0.04 Hz) reflects body temperature regulation and vascular tone; LF band (0.04-0.15 Hz) reflects the combined activity of the sympathetic and parasympathetic nervous systems; HF band (0.15-0.4 Hz) mainly reflects parasympathetic activity. The mobile device calculates the power values ​​of each frequency band and performs logarithmic transformation.

[0067] (4) Frequency domain index calculation: Calculate the LF / HF ratio to reflect the balance between the sympathetic and parasympathetic nervous systems; calculate the total power (TP) to reflect the overall activity level of the autonomic nervous system; calculate the normalized power to eliminate the influence of individual differences. Mobile devices standardize the frequency domain indexes to improve comparability.

[0068] refer to Figure 2 As shown, in some embodiments of this specification, the physiological state features may include at least nonlinear features; based on the current physiological signal, extracting multiple dimensions of physiological state features may include: S201: Match the R-wave interval data based on a preset sample template to calculate the sample entropy of the R-wave interval data.

[0069] It is understandable that sample entropy is an indicator for measuring the complexity and regularity of a time series. The larger the value, the more complex the series is, and the better it reflects the function of the autonomic nervous system.

[0070] In practice, the template length can be set to m=2, and the tolerance to r=0.2×SD (SD is the standard deviation of the RR interval data, 45ms, i.e., r=9ms); two m-dimensional vectors X can be defined. i =(RR i , RR i+1 ) and X j =(RR j , RR j+1 (i≠j); calculate the distance d(X) between vectors. i Xj )=max (|RR i+k - RR j+k |)(k=0,1;Statistical analysis of d(X) i X j The proportion of vector pairs less than r to the total number of vector pairs is denoted as C. i m (r); Calculate C m (r)=(1 / (N - m))ΣC i m (r) (N is the number of RR interval data, 300); Sample entropy SampEn=-ln [C m+ ¹(r) / C m (r)].

[0071] S202: Based on the R-wave interval data, the detrended fluctuation function corresponding to the R-wave interval data is calculated using the piecewise integration method, and the scaling exponent of the detrended fluctuation function is calculated as the self-similarity of the R-wave interval data.

[0072] It can be understood that self-similarity is the scaling index α obtained through detrended fluctuation analysis (DFA). An α value close to 1.0 indicates that the heart rate signal has healthy self-similarity characteristics.

[0073] In practice, the RR interval series can be divided into N non-overlapping intervals of length n; a linear fit can be performed on each interval to obtain a trend line; and the deviation of each data point from the trend line can be calculated to obtain the detrending oscillation function. ; Calculate the double logarithmic relationship between F(n) and n to obtain the scaling exponent α.

[0074] S203: Using the ellipse fitting method to generate a Poincaré diagram corresponding to the R-wave interval data, calculate the geometric features of the Poincaré diagram, wherein the geometric features include at least one of the following: the major axis, minor axis, and area of ​​the ellipse.

[0075] In practice, the geometric features of the Poincaré diagram can be represented by RR. i For the horizontal axis, RR i+1 The geometric parameters (major axis, minor axis, area) of the scatter plot plotted with the vertical axis reflect the nonlinear dynamic characteristics of the RR interval series.

[0076] S204: Calculate the first complexity of the R-wave interval data at different time scales, the permutation entropy of the R-wave interval data, and the second complexity of the R-wave interval sequence corresponding to the R-wave interval data, and use the first complexity, the permutation entropy, and the second complexity as the complexity of the R-wave interval data.

[0077] Complexity is understood to be an indicator of the randomness and regularity of RR interval sequences, including multi-scale entropy, permutation entropy, and Lempel-Ziv complexity. The first type of complexity is multi-scale entropy, which measures the complexity of the RR interval sequence at different time scales. Permutation entropy measures the randomness of a time series. The second type of complexity is Lempel-Ziv complexity, which assesses the regularity of the sequence. These complexities can provide multiple dimensions of heart rate variability.

[0078] In some embodiments of this specification, determining the current state of the target object based on the multiple dimensions of physiological state characteristics may include: determining the feature weights of the physiological state characteristics of each dimension, and calculating the current stress index of the target object as the current state based on the physiological state characteristics of each dimension and the feature weights of each dimension.

[0079] In the embodiments of this specification, by extracting nonlinear features that reflect the complex dynamic characteristics of heart rate signals, the potential state of the autonomic nervous system can be further revealed, improving the depth and accuracy of physiological state assessment. Through the combined use of indicators such as multi-scale entropy, permutation entropy, and Lempel-Ziv complexity, the complexity and regularity of the RR interval sequence are characterized from different perspectives, providing a more comprehensive reflection of the user's physiological state changes and offering richer evidence for subsequent state judgment and audio adjustment. The method of generating Poincaré diagrams using ellipse fitting and extracting geometric features intuitively and quantitatively reflects the nonlinear characteristics of heart rate variability. Compared with traditional qualitative analysis, this improves the interpretability and application value of the features, enabling audio adjustment to more accurately adapt to the user's physiological state.

[0080] In some embodiments of this specification, the physiological state characteristics of each dimension may include index values ​​corresponding to at least two physiological state indicators. Accordingly, determining the feature weights of the physiological state characteristics of each dimension, and calculating the current stress index of the target object as the current state based on the physiological state characteristics and feature weights of each dimension, may include: determining the index weights corresponding to each physiological state indicator, and calculating the current stress index as the current state based on the index values ​​and index weights of each physiological state indicator.

[0081] As can be understood, feature weights refer to the importance coefficients of each dimension of physiological state features in the calculation of the stress index. The larger the weight value, the greater the influence of that dimension feature on the stress state. Feature weights can be preset in the system based on physiological theories, clinical trial data, and machine learning optimization.

[0082] In practice, the calculation of the current pressure index may include: S1, Weighting coefficient setting.

[0083] For example, based on physiological research and clinical experience, the system sets weight coefficients for each indicator. RMSSD weight 0.25: reflects parasympathetic neural activity. LF / HF ratio weight 0.30: reflects sympathetic / parasympathetic balance. SDNN weight 0.20: reflects overall variability. pNN50 weight 0.15: reflects high-frequency variability. Sample entropy weight 0.10: reflects complexity.

[0084] S2, index normalization processing.

[0085] For example, mobile devices normalize each HRV indicator to eliminate the influence of dimensions; the min-max normalization method is used to map the indicator values ​​to the 0-1 range; for positive indicators such as RMSSD, the larger the value, the less pressure; for negative indicators such as LF / HF ratio, the larger the value, the greater the pressure.

[0086] S3, Calculation of Comprehensive Pressure Index.

[0087] For example, the mobile device system uses a weighted average method to calculate the overall stress index. Stress index = Σ(weight i × normalized index i); the stress index ranges from 0 to 1, with a larger value indicating a higher stress level. The mobile device smooths the stress index to reduce short-term fluctuations.

[0088] In some embodiments of this specification, the above method may further include: obtaining a pressure index baseline corresponding to the target object, the pressure index baseline including multiple preset pressure states and pressure index intervals corresponding to each pressure state; and determining the current pressure state of the target object as the current state based on the current pressure index and the pressure index baseline.

[0089] It is understandable that the stress index baseline refers to a personalized stress status assessment standard established for a specific target object, including a preset stress status category (such as low stress, moderate stress, high stress) and the corresponding stress index range for each category. The establishment of the stress index baseline is based on the target object's historical physiological data to ensure the personalization and accuracy of stress status judgment.

[0090] refer to Figure 3 As shown, in some embodiments of this specification, obtaining the baseline pressure index corresponding to the target object may include: S301: Select a portion of the historical physiological signals from the target object at multiple times that meet a preset time threshold as the target physiological signal set.

[0091] S302: Calculate the historical stress index corresponding to each historical physiological signal in the target physiological signal set, and determine the distribution characteristics of the historical stress index.

[0092] S303: Based on the distribution characteristics, determine the pressure index range corresponding to each pressure state.

[0093] It is understandable that a preset time threshold is used to filter the time range of valid historical physiological signals, set to 30 days (balancing data timeliness and sample size), meaning only historical physiological signals within 30 days of the current time are selected. The target physiological signal set is the set of valid historical physiological signals after filtering, requiring a sample size of ≥30 groups (each group containing 5-minute RR interval data) to ensure the reliability of the distribution characteristics statistics. The distribution characteristics are the statistical distribution attributes of the historical stress index, including percentiles (33rd and 67th percentiles), mean, standard deviation, and distribution density. Using percentiles to divide the intervals conforms to the natural distribution law of stress states.

[0094] In practice, during the initial user phase, mobile devices can guide users to perform multiple short-duration (e.g., 5-minute) heart rate measurements in a relaxed state (e.g., 3 times daily for 7 days), calculating the PI value for each measurement to form an initial baseline dataset. The system periodically (e.g., every two weeks) updates this baseline using all PI data within a sliding window (e.g., the most recent 30 days). Based on the distribution of this dataset, the system dynamically calculates classification thresholds; for example, the threshold for low-stress states is the 33rd percentile (P33), the threshold for high-stress states is the 67th percentile (P67), and states in between are considered moderate-stress states. This method, based on individual historical percentiles, automatically adapts to the inherent differences in HRV levels among different users.

[0095] In some embodiments of this specification, the pressure index baseline further includes physiological state feature intervals corresponding to each pressure state; the method may further include: determining an initial pressure state corresponding to the current pressure index based on the pressure index baseline; when determining that the initial pressure state is in the boundary region of a pressure state category, determining a reference pressure state corresponding to the physiological state feature based on the pressure index baseline; and determining the current pressure state of the target object as the current state based on the reference pressure state and the initial pressure state.

[0096] It is understandable that, to further improve the accuracy of determining the current stress state, the stress state can be corrected by combining the aforementioned determined physiological state characteristics. The physiological state characteristic range is a preset range of values ​​for each physiological state characteristic within the stress index baseline for each stress state. It serves as the basis for judging the matching degree between the physiological state characteristics and the stress state, including the numerical ranges of time-domain, frequency-domain, and nonlinear characteristics (e.g., the RMSSD characteristic range for low-pressure states is 50-80ms). The boundary region is the critical range between two adjacent stress state ranges, set as an interval of ±5% of the stress state threshold (e.g., the boundary region corresponding to the low-pressure upper limit of 0.3 is 0.285-0.315, and the boundary regions corresponding to the medium-pressure upper and lower limits of 0.3 and 0.7 are 0.285-0.315 and 0.665-0.735, respectively), used to identify ambiguous stress states. The reference stress state is the stress state that best matches the current characteristics by comparing the current physiological state characteristics with the physiological characteristic ranges of each state in the stress index baseline, used to assist in judging the stress state in the boundary region.

[0097] In practice, the mobile device can determine an initial stress state based on the dynamic threshold range into which the current PI value falls. Then, the system performs a second level of verification: checking if the current PI value is near the classification boundary (e.g., within 5% of the P33 or P67 threshold). If so, an arbitration mechanism is initiated. The arbitration is based on whether key physiological indicators (such as RMSSD and LF / HF) simultaneously support the classification. For example, if the initial classification is moderate stress, but the current RMSSD is extremely low and the LF / HF is extremely high (both strongly indicating high stress), the system may correct the state to high stress. For more complex boundary cases, the system can use fuzzy logic to calculate the membership degree of the current feature to each stress state, and use the state with the highest membership degree as the final current state.

[0098] In the embodiments described in this specification, a high degree of personalization and robustness of stress state assessment is achieved by using methods such as weighted calculation of comprehensive index, setting dynamic thresholds based on individual historical percentiles, and multi-level classification and key indicator arbitration. This can solve the problem that general fixed thresholds are not suitable for individual differences, and intelligent arbitration reduces classification misjudgments caused by short-term data fluctuations, thereby providing extremely reliable and personalized input for subsequent audio adjustment, significantly improving the adaptability and effectiveness of the entire system.

[0099] In some embodiments of this specification, the current stress index can be classified using a stress state classifier to obtain the current state. The stress state classifier can preset the following stress states: low stress state (stress index 0-0.3, good HRV); medium stress state (stress index 0.3-0.7, medium HRV); high stress state (stress index 0.7-1.0, poor HRV). The mobile device can dynamically adjust the classification threshold based on user baseline data.

[0100] Furthermore, after determining the classification results, these results can be verified. Mobile devices can perform consistency checks on the classification results; verify the accuracy of the classification by combining multiple HRV indicators; trigger re-analysis for abnormal classification results; and maintain statistical information on classification accuracy on mobile devices.

[0101] In some embodiments of this specification, dynamically adjusting at least one audio parameter of the audio to be played may include at least one of the following: adjusting the harmonic harmony of the audio to be played; adjusting the volume and / or frequency of each sound element of the audio to be played; adjusting the rhythm parameter of the audio to be played, wherein the rhythm parameter includes at least one of the following: beat frequency and beat interval duration.

[0102] Harmony can be understood as a quantitative indicator (range 0-1) that measures the degree of coordination in an audio chord sequence. A higher value indicates a more relaxed and tension-free chord combination, calculated based on chord type, function, and connection relationships. Sound elements are the basic units that constitute audio, including musical instruments (such as guitar, piano, drums), spectral components (high frequency, mid frequency, low frequency), and sound effects (distortion, reverberation), etc. Rhythm parameters are the core indicators that determine the rhythmic characteristics of audio. Beat frequency (BPM) refers to the number of beats per minute, and beat interval refers to the time difference between two adjacent beats (such as the interval between quarter notes in 4 / 4 time).

[0103] refer to Figure 4 As shown, in some embodiments of this specification, adjusting the harmonic harmony of the audio to be played may include: S401: Determine the harmonic structure of the audio to be played, wherein the harmonic structure includes a chord sequence composed of multiple chord elements and the chord parameters of each chord element.

[0104] S402: Based on the mapping relationship between chords and scores, the harmonic weights corresponding to each chord, and the harmonic structure, determine the harmonic harmony of the audio to be played.

[0105] S403: Determine the target harmony degree corresponding to the current state.

[0106] S404: Based on the difference between the harmonic harmony and the target harmony, a progressive conversion is used to adjust the harmonic structure of the audio to be played.

[0107] It is understandable that chord elements can form the basic units of harmonic structure. The chord parameters of each chord element can include chord type (major triad, minor triad, diminished triad, augmented triad, seventh chord, etc.), chord function (tonic chord, subdominant chord, dominant chord, secondary chord), chord key (e.g., C major, G major), and connection type (stepwise, leap). The mapping relationship between chords and scores can be a preset correspondence between chord types and basic harmony scores (e.g., major triad = 1.0, minor triad = 0.8, dominant seventh chord = 0.6, diminished triad = 0.3), stored in the system database. Harmonic weights can be weight coefficients set based on the importance of chord functions (e.g., tonic chord = 1.0, subdominant chord = 0.8, dominant chord = 0.6, secondary chord = 0.4). Progressive transitions are a way to replace dissonant chords in stages without disrupting the smoothness of the audio, usually in units of 2-4 measures, to avoid auditory discomfort caused by abrupt changes in chords.

[0108] In some embodiments of this specification, determining the harmonic harmony of the audio to be played, based on the mapping relationship between each harmony and the score, the harmonic weights corresponding to each harmony, and the harmonic structure, may include: obtaining the functional weights of each chord element in the chord sequence; calculating the context weights of each chord element based on the chord sequence and the chord parameters of each chord element; determining the score corresponding to each chord element based on the mapping relationship; and calculating the harmonic harmony of the audio to be played based on the score corresponding to each chord element, the functional weights of each chord element, and the context weights of each chord element.

[0109] In practice, the adjustment of harmonic harmony can include: (1) Chord basic harmony score.

[0110] Mobile devices can establish a chord harmony rating database (i.e., the mapping relationship mentioned above), covering various chord types. For example: major triad rating 1.0: the most harmonious chord type; minor triad rating 0.8: a relatively harmonious chord type; major seventh chord rating 0.9: a seventh chord with high harmony; minor seventh chord rating 0.7: a seventh chord with medium harmony; dominant seventh chord rating 0.6: a chord with some tension; diminished triad rating 0.3: a dissonant chord type; augmented triad rating 0.2: a highly dissonant chord; diminished seventh chord rating 0.1: the most dissonant chord type.

[0111] (2) Chord function weight setting.

[0112] For example, the tonic chord weight is 1.0: the most stable chord function; the subdominant chord weight is 0.8: a relatively stable chord function; the dominant chord weight is 0.6: a chord function with a certain tension; and the secondary chord weight is 0.4: a chord function with greater tension. Mobile devices set the chord function weights according to music theory.

[0113] (3) Context weight calculation.

[0114] Mobile devices analyze the relationship between the current chord and the preceding and following chords; calculate the smoothness and musicality of chord transitions; consider tonal relationships and harmonic progression rules; and adjust context weights according to musical style.

[0115] (4) Comprehensive assessment of harmony.

[0116] Mobile devices use a weighted average method to calculate the overall harmony score. Harmony score = Σ(base score × functional weight × context weight) / number of chords. Mobile devices normalize the harmony score, which ranges from 0 to 1. The higher the harmony score, the more harmonious the music is and the more conducive it is to relaxation.

[0117] (5) Chord adjustment algorithm, including target harmony setting, current harmony evaluation, chord conversion strategy, conversion path planning, and real-time chord adjustment.

[0118] The target harmony level is set as follows: under low pressure, the target harmony level is 0.9, maintaining high harmony; under medium pressure, the target harmony level is 0.7, achieving moderate harmony; and under high pressure, the target harmony level is 0.5, gradually increasing harmony. The mobile device dynamically adjusts the target harmony level according to the pressure state.

[0119] The current harmony assessment includes: real-time analysis of the chord progressions of the currently playing music by the mobile device; identification of the type, function, and contextual relationship of each chord; calculation of the overall harmony of the current music; and comparison with the target harmony to determine adjustment needs.

[0120] Chord transition strategies include: mobile devices identifying dissonant chords and developing transition plans; converting diminished triads to minor triads to reduce dissonance; converting augmented triads to major triads to increase harmony; converting diminished seventh chords to minor seventh chords to improve harmonic effects; and using gradual transitions on mobile devices to avoid abrupt changes.

[0121] The conversion path planning includes: mobile devices planning the optimal conversion path based on music theory; considering chord function relationships and tonal logic; ensuring the musicality and smoothness of the conversion process; and mobile devices supporting multiple conversion strategies to adapt to different music styles.

[0122] Real-time chord adjustment includes: real-time monitoring of music playback progress on mobile devices; applying chord transitions at appropriate times; achieving seamless transitions using audio processing technology; and maintaining the continuity and integrity of the music.

[0123] In some embodiments of this specification, adjusting the volume and / or frequency of each sound element of the audio to be played may include at least one of the following: The instruments contained in the audio to be played are identified as the first sound elements; the instrument index corresponding to each instrument is obtained, and the instrument index is used to characterize the degree of physiological influence of the corresponding instrument on the target object; the volume of each instrument is adjusted based on the mapping relationship between stress state and instrument index, combined with preset volume adjustment rules. The spectrum of the audio to be played is determined as the second sound element. Based on the mapping relationship between the second sound element, pressure state and frequency band gain value, the frequency band gain value of the audio to be played is adjusted. The spectral characteristics of each instrument in the audio to be played are determined as the third sound element. Based on the mapping relationship between the third sound element, the pressure state and the frequency band gain value of each instrument, the frequency band gain value of each instrument is adjusted.

[0124] The instrument index is a quantitative indicator (range 0-1) representing the degree of physiological impact of an instrument on the user. It is divided into a stimulation index (higher values ​​indicate a greater likelihood of inducing tension) and a soothing index (higher values ​​indicate a greater likelihood of relieving stress), which are negatively correlated (stimulation index = 1 - soothing index), and are preset in the instrument database. The preset volume adjustment rule is a volume adjustment logic based on stress level and instrument index. The higher the stress, the greater the volume attenuation for highly stimulating instruments and the greater the volume enhancement for highly soothing instruments. The frequency band gain value is the volume gain parameter (in dB) for each frequency band of the audio (high frequency 2-8kHz, mid frequency 500Hz-2kHz, low frequency 20-500Hz). Negative values ​​indicate attenuation, and positive values ​​indicate enhancement. The spectral characteristics of the instruments are the energy distribution of each instrument in different frequency bands (e.g., the high-frequency energy of the electric guitar is concentrated in 3-6kHz, and the low-frequency energy of the cello is concentrated in 100-300Hz).

[0125] In practice, adjusting the first sound element (instrument volume) may include: identifying instruments in the audio (using spectral analysis and an instrument recognition model) to obtain a set of instruments: electric guitar, acoustic guitar, piano, cymbals, and cello. Obtaining the instrument index for each instrument: electric guitar (stimulation index 0.9), acoustic guitar (soothing index 0.3), piano (soothing index 0.2), cymbals (stimulation index 0.8), and cello (soothing index 0.3). Determining the current stress state (high stress, index 0.75), and the mapping relationship between stress state and instrument index: under high stress, instruments with a stimulation index > 0.7 have their volume reduced by 30-50%, while instruments with a soothing index > 0.2 have their volume increased by 20-30%. Performing volume adjustments: electric guitar volume reduced by 40%, cymbal volume reduced by 50%, acoustic guitar volume increased by 25%, cello volume increased by 30%, and piano volume increased by 20%.

[0126] In practice, adjusting the second sound element (spectral gain) may include: using FFT to analyze the audio spectrum to obtain the energy distribution of each frequency band: high frequency (2-8kHz) energy proportion, mid frequency (500Hz-2kHz) proportion, and low frequency (20-500Hz) proportion. Obtain the mapping relationship between pressure state and frequency band gain values: high pressure state → high frequency -3dB, mid frequency -1dB, low frequency +2dB; medium pressure → high frequency -1dB, mid frequency 0dB, low frequency 0dB; low pressure → high frequency -2dB, low frequency +1dB. Given a high pressure state, adjust the frequency band gain values: high frequency attenuation 3dB, mid frequency attenuation 1dB, and low frequency enhancement 2dB.

[0127] In practice, adjusting the third sound element (instrument frequency band gain) may include: extracting the spectral characteristics of each instrument, including electric guitar (3-6kHz energy focus), acoustic guitar (1-3kHz energy focus), and cello (100-300Hz energy focus). Obtaining the mapping relationship between stress levels and the gain values ​​of each instrument's frequency bands: under high stress, the 3-6kHz frequency band of the electric guitar is attenuated by 4dB, the 1-3kHz frequency band of the acoustic guitar is boosted by 2dB, and the 100-300Hz frequency band of the cello is boosted by 3dB. Performing instrument frequency band gain adjustment: the 3-6kHz frequency band of the electric guitar is attenuated by 4dB (reducing sharpness), the 1-3kHz frequency band of the acoustic guitar is boosted by 2dB (emphasizing a warm tone), and the 100-300Hz frequency band of the cello is boosted by 3dB (enhancing the soothing effect of low frequencies).

[0128] In the embodiments of this specification, sound elements are adjusted from three levels: overall instrument volume, global audio spectrum, and local instrument frequency bands. This achieves fine-grained adjustment, which, compared to the overall EQ adjustment of existing technologies, better suits the acoustic characteristics and stress relief needs of different instruments. The introduction of the instrument index establishes a direct link between the instrument and the physiological state, making the adjustment logic more physiologically based and avoiding blind adjustment. It supports the combined use of three adjustment methods (such as simultaneously adjusting instrument volume and frequency band gain), which can be flexibly selected according to the audio type (instrumental music, song) and user status, resulting in more significant adjustment effects. The preset mapping relationship and adjustment rules can be continuously optimized through user feedback, enhancing the personalization and adaptability of the embodiments of this specification.

[0129] In some embodiments of this specification, adjusting the rhythm parameters of the audio to be played may include: S1. Determine multiple beat points contained in the audio to be played, and determine the beat frequency of the audio to be played based on the multiple beat points; S2. Obtain the target beat frequency range corresponding to the current state; adjust the beat frequency based on the target beat frequency range, the beat frequency, and a preset beat frequency adjustment rule; and / or, S3. Calculate the beat interval duration for each pair of beat points; determine the standard deviation and coefficient of variation for multiple beat points based on the beat interval duration; adjust at least one beat interval duration based on the difference between the beat frequency and the target beat frequency range; and / or, S4. Obtain the beat heart rate synchronization rule corresponding to the current state, and adjust the beat frequency based on the beat heart rate synchronization rule.

[0130] It is understood that, in the embodiments of this specification, when adjusting the rhythm parameters, after determining the beat point and beat frequency, at least one of the above steps S2 to S4 can be performed to adjust the rhythm parameters.

[0131] It can be understood that beat points are the time points in audio that have strong or weak beat characteristics (such as the moments corresponding to drum beats or bass accents), identified through the onset detection algorithm. The preset beat frequency adjustment rule determines the adjustment range based on the difference between the current BPM and the target range; the greater the difference, the larger the adjustment range, but a single adjustment cannot exceed 20 BPM (to avoid sudden rhythmic changes). The PSOLA algorithm is used to ensure that the pitch remains unchanged after adjustment. The standard deviation and coefficient of variation of the beat interval duration reflect the stability of the beat interval; the smaller the standard deviation and coefficient of variation, the smoother the rhythm (coefficient of variation = standard deviation / mean). The beat heart rate synchronization rule is a rule for synchronizing the music beat frequency with the user's current heart rate. For example, it can be to make the music BPM close to 0.8-1.2 times the user's resting heart rate, enhancing the coordination between physiology and music.

[0132] In practice, adjusting the rhythm parameters may include the following steps: S1. Target BPM range setting: Low stress state: BPM range 60-80, soothing rhythm; Medium stress state: BPM range 80-100, medium rhythm; High stress state: BPM range 60-70, forced soothing; The mobile device dynamically adjusts the target BPM according to the stress state.

[0133] S2. Beat Detection: Mobile devices use audio analysis technology to detect music beats. The onset detection algorithm can be used to identify beat points and calculate the stability and regularity of the beats. Multiple beat modes are supported: 4 / 4, 3 / 4, 6 / 8, etc.

[0134] S3, BPM Adjustment: The mobile device detects the current music's BPM value; compares it with the target BPM range; forcibly reduces the BPM under high pressure; and uses time stretching technology to maintain a constant pitch.

[0135] The time stretching technology can include: mobile devices using the PSOLA (Pitch Synchronous Overlap and Add) algorithm; changing the playback speed while keeping the pitch constant; ensuring that the audio quality is not affected; and mobile devices supporting real-time time stretching processing.

[0136] S4. Enhanced Beat Stability: The mobile device analyzes changes in beat intervals; calculates the standard deviation and coefficient of variation of the beat intervals; and improves stability by fine-tuning the beat intervals. The mobile device can employ adaptive algorithms to optimize beat patterns.

[0137] S5, Rhythm Synchronization: Mobile devices analyze the user's heart rate change patterns; synchronize the music beat with the user's heart rate; enhance the coordination between music and physiology, and can support multiple synchronization modes.

[0138] The embodiments in this specification cover multiple dimensions of rhythm adjustment (beat frequency, interval stability, and heart rate synchronization), achieving refined and multi-dimensional rhythm adjustment, which better meets the physiological needs of stress relief. The introduction of beat-heart rate synchronization rules establishes a direct correlation between music rhythm and user physiological signals, enhancing the coordination between music and physiological state, and making the adjustment effect more significant. It supports flexible combinations of multiple adjustment methods, and the optimal adjustment strategy can be selected according to the audio type (rhythm type, melody type) and user state, enhancing practicality and adaptability.

[0139] In some embodiments of this specification, audio processing can be implemented using a real-time audio processing engine. The real-time audio processing engine may include: (1) Audio stream processing architecture: A multi-threaded architecture is used to process audio streams; the main thread is responsible for audio playback control; the processing thread is responsible for audio parameter adjustment; and the analysis thread is responsible for music feature analysis.

[0140] (2) Processing pipeline: Establish a modular audio processing pipeline, with each module responsible for specific processing functions. Support dynamic addition and removal of processing modules, and use standard interfaces to communicate between modules.

[0141] (3) Real-time parameter adjustment: Real-time monitoring of music playback progress; applying parameter adjustment at appropriate times; using crossfade technology to achieve seamless switching; maintaining audio continuity and smoothness.

[0142] (4) Audio quality assurance: High-quality audio processing algorithms are adopted, supporting multiple audio formats such as WAV, MP3, and FLAC, maintaining the dynamic range and sound quality of the original audio, and supporting real-time sound quality monitoring.

[0143] (5) Delay control module: optimizes processing delay and controls it within 100ms; uses prediction algorithm to preprocess audio data; uses buffering technology to balance delay and quality; supports delay compensation mechanism.

[0144] Furthermore, the audio adjustment process can be optimized based on a feedback control mechanism. This feedback control mechanism specifically includes the following: (1) Evaluation of the effect of regulation: monitor the trend of HRV index; calculate the decrease of stress index; evaluate the effectiveness of music regulation; and establish an effect evaluation model.

[0145] (2) Parameter adaptive adjustment: Adjust the parameter intensity according to the adjustment effect; increase the adjustment intensity when the effect is not good; decrease the adjustment intensity when the effect is too strong; use PID control algorithm to optimize parameters.

[0146] (3) Historical data management: Maintain users' historical data, including HRV indicators, stress status, adjustment parameters, etc., support data analysis and pattern recognition, and regularly clean up expired data.

[0147] (4) Learning optimization mechanism: learn the user's physiological response pattern, optimize personalized regulation strategy, establish user feature model, and support online learning and updating.

[0148] (5) Anomaly handling mechanism: detect abnormal situations and trigger the handling mechanism, including connection interruption, data anomaly, processing error, etc., provide automatic recovery and manual intervention options, and maintain anomaly log and statistical information.

[0149] In the embodiments of this specification, a deep HRV analysis system is established using the above methods. This system comprehensively assesses the user's autonomic nervous system state using multiple dimensions of HRV indicators (time domain, frequency domain, and nonlinearity). By establishing a quantitative model of musical harmony, dynamic adjustment of the intrinsic structure of music, such as chord progressions, orchestration, and rhythm, is achieved. By establishing a mapping relationship between HRV indicators and musical parameters, closed-loop feedback regulation is achieved, thereby constructing a deep interaction mechanism between biological signals and music. Music adjustment is automatically performed based on objective physiological indicators without user intervention, achieving active stress regulation. Personalized music adjustment is provided according to the user's physiological characteristics and stress patterns, achieving personalized music therapy.

[0150] Based on the above-described audio modulation method based on physiological signals, one or more embodiments of this specification also provide an audio modulation device based on physiological signals. The device may include an apparatus (including a distributed system), software (application), module, plug-in, server, client, etc., using the method described in the embodiments of this specification, combined with necessary implementation hardware. Based on the same innovative concept, the devices in one or more embodiments provided in this specification are as described in the following embodiments. Since the implementation schemes and methods for solving the problem are similar, the implementation of specific devices in the embodiments of this specification can refer to the implementation of the foregoing method, and repeated details will not be elaborated further. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated. Figure 5 The diagram shown is a schematic representation of an audio modulation device based on physiological signals provided in an embodiment of this application. Figure 5 As shown, the audio modulation device 500 based on physiological signals may include: Acquisition module 501 is used to acquire the current physiological signals of the target object; Extraction module 502 is used to extract physiological state features of multiple dimensions based on the current physiological signal; The determining module 503 is used to determine the current state of the target object based on the multiple dimensions of physiological state characteristics; The adjustment module 504 is used to dynamically adjust at least one audio parameter of the audio to be played according to the current state, so as to generate and output the audio to be played that matches the current state.

[0151] In some embodiments of this specification, the current physiological signal includes R-wave interval data, and the physiological state characteristics include at least two of the following: time-domain characteristics, frequency-domain characteristics, and nonlinear characteristics; the time-domain characteristics include at least one of the following: statistical characteristics of R-wave interval data, root mean square of the difference between adjacent R-wave intervals, proportion of adjacent R-wave interval differences greater than a preset threshold, and geometric characteristics of the distribution of adjacent R-wave intervals; the frequency-domain characteristics include at least one of the following: frequency band power of the frequency domain signal corresponding to the R-wave interval data in multiple preset frequency bands, total power of the frequency domain signal, normalized power of the frequency domain signal, and ratio of power of at least two frequency bands; the nonlinear characteristics include at least one of the following: sample entropy of R-wave interval data, self-similarity of R-wave interval data, geometric characteristics of the Poincaré diagram corresponding to the R-wave interval data, and complexity of the R-wave interval data.

[0152] In some embodiments of this specification, the physiological state features include at least nonlinear features; the extraction module 502 is specifically used for: matching the R-wave interval data based on a preset sample template to calculate the sample entropy of the R-wave interval data; calculating the detrended fluctuation function corresponding to the R-wave interval data using a piecewise integration method based on the R-wave interval data, and calculating the scaling exponent of the detrended fluctuation function as the self-similarity of the R-wave interval data; generating a Poincaré diagram corresponding to the R-wave interval data using an ellipse fitting method, and calculating the geometric features of the Poincaré diagram, wherein the geometric features include at least one of the following: the major axis, minor axis, and area of ​​the ellipse; calculating the first complexity of the R-wave interval data at different time scales, the permutation entropy of the R-wave interval data, and the second complexity of the R-wave interval sequence corresponding to the R-wave interval data, and using the first complexity, the permutation entropy, and the second complexity as the complexity of the R-wave interval data.

[0153] In some embodiments of this specification, the determining module 503 is specifically used to: determine the feature weights of physiological state features in each dimension, and calculate the current stress index of the target object as the current state based on the physiological state features in each dimension and the feature weights of each dimension.

[0154] In some embodiments of this specification, the physiological state characteristics of each dimension include index values ​​corresponding to at least two physiological state indicators; accordingly, when determining the feature weights of the physiological state characteristics of each dimension and calculating the current stress index of the target object as the current state based on the physiological state characteristics and feature weights of each dimension, the determining module 503 is specifically used to: determine the index weights corresponding to each physiological state indicator, and calculate the current stress index as the current state based on the index values ​​and index weights of each physiological state indicator.

[0155] In some embodiments of this specification, the determining module 503 may also be used to: obtain a pressure index baseline corresponding to the target object, the pressure index baseline including multiple preset pressure states and pressure index intervals corresponding to each pressure state; and determine the current pressure state of the target object as the current state based on the current pressure index and the pressure index baseline.

[0156] In some embodiments of this specification, the pressure index baseline further includes physiological state feature intervals corresponding to each pressure state; the determination module 503 can also be used to: determine the initial pressure state corresponding to the current pressure index based on the pressure index baseline; when it is determined that the initial pressure state is in the pressure state category boundary region, determine the reference pressure state corresponding to the physiological state feature based on the pressure index baseline; and determine the current pressure state of the target object as the current state based on the reference pressure state and the initial pressure state.

[0157] In some embodiments of this specification, the acquisition module 501 is specifically used to: select a portion of the historical physiological signals that meet a preset time threshold from the current time as the target physiological signal set from the historical physiological signals of the target object at multiple times; calculate the historical pressure index corresponding to each historical physiological signal in the target physiological signal set, and determine the distribution characteristics of the historical pressure index; and determine the pressure index interval corresponding to each pressure state based on the distribution characteristics.

[0158] In some embodiments of this specification, the adjustment module 504 dynamically adjusts at least one audio parameter of the audio to be played, for at least one of the following: adjusting the harmonic harmony of the audio to be played; adjusting the volume and / or frequency of each sound element of the audio to be played; adjusting the rhythm parameters of the audio to be played, the rhythm parameters including at least one of the following: beat frequency, beat interval duration.

[0159] In some embodiments of this specification, the adjustment module 504 adjusts the harmonic harmony of the audio to be played, specifically for: determining the harmonic structure of the audio to be played, the harmonic structure including a chord sequence composed of multiple chord elements and chord parameters of each chord element; determining the harmonic harmony of the audio to be played based on the mapping relationship between chords and scores, the harmonic weights corresponding to each chord, and the harmonic structure; determining the target harmony corresponding to the current state; and adjusting the harmonic structure of the audio to be played using a progressive conversion based on the difference between the harmonic harmony and the target harmony.

[0160] In some embodiments of this specification, when the adjustment module 504 determines the harmonic harmony of the audio to be played based on the mapping relationship between each harmony and the score, the harmonic weights corresponding to each harmony, and the harmonic structure, it is specifically used to: obtain the functional weights of each chord element in the chord sequence; calculate the context weights of each chord element based on the chord sequence and the chord parameters of each chord element; determine the score corresponding to each chord element based on the mapping relationship; and calculate the harmonic harmony of the audio to be played based on the score corresponding to each chord element, the functional weights of each chord element, and the context weights of each chord element.

[0161] In some embodiments of this specification, when the adjustment module 504 adjusts the volume and / or frequency of each sound element of the audio to be played, it is specifically used to: determine the musical instruments included in the audio to be played as the first sound element, obtain the instrument index corresponding to each instrument, the instrument index being used to characterize the degree of physiological influence of the corresponding instrument on the target object, and adjust the volume of each instrument based on the mapping relationship between the pressure state and the instrument index, combined with a preset volume adjustment rule; and / or, determine the spectrum of the audio to be played as the second sound element, and adjust the frequency band gain value of the audio to be played based on the mapping relationship between the second sound element, the pressure state, and the frequency band gain value; and / or, determine the spectral characteristics of each instrument included in the audio to be played as the third sound element, and adjust the frequency band gain value of each instrument based on the mapping relationship between the third sound element, the pressure state, and the frequency band gain value of each instrument.

[0162] In some embodiments of this specification, when the adjustment module 504 adjusts the rhythm parameters of the audio to be played, it is specifically used to: determine multiple beat points contained in the audio to be played, and determine the beat frequency of the audio to be played based on the multiple beat points; obtain a target beat frequency range corresponding to the current state, and adjust the beat frequency based on the target beat frequency range, the beat frequency, and a preset beat frequency adjustment rule; and / or, calculate the beat interval duration between pairs of beat points, determine the standard deviation and coefficient of variation of multiple beat points based on the beat interval duration, and adjust at least one beat interval duration based on the difference between the beat frequency and the target beat frequency range; and / or, obtain a beat heart rate synchronization rule corresponding to the current state, and adjust the beat frequency based on the beat heart rate synchronization rule.

[0163] The descriptions and functions of the above modules can be found in the section on audio modulation methods based on physiological signals, and will not be repeated here.

[0164] This application also provides an electronic device, such as... Figure 6As shown, the electronic device may include a processor 601 and a memory 602, wherein the processor 601 and the memory 602 may be connected via a bus or other means. Figure 6 Taking the example of a connection between China and Israel via a bus.

[0165] Processor 601 may be a central processing unit (CPU). Processor 601 may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations thereof.

[0166] The memory 602, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the audio modulation method based on physiological signals in the embodiments of the present invention. The processor 601 executes various functional applications and data processing of the processor by running the non-transitory software programs, instructions, and modules stored in the memory 602, thereby realizing the audio modulation method based on physiological signals in the above method embodiments.

[0167] The memory 602 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created by the processor 601, etc. Furthermore, the memory 602 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 602 may optionally include memory remotely located relative to the processor 601, and these remote memories may be connected to the processor 601 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0168] The one or more modules are stored in the memory 602 and are executed by the processor 601. Figure 1 The audio modulation method based on physiological signals described in the embodiments.

[0169] The specific details of the aforementioned electronic device can be understood by referring to the relevant descriptions and effects in the above method embodiments, and will not be repeated here.

[0170] This specification also provides a computer storage medium storing computer program instructions that, when executed, implement the steps of the above-described audio modulation method based on physiological signals.

[0171] This specification also provides a computer program product comprising a computer program that, when executed, implements the steps of the above-described audio modulation method based on physiological signals.

[0172] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium can also include combinations of the above types of memory.

[0173] The various embodiments in this specification are described in a progressive manner. For the same or similar parts between the various embodiments, please refer to each other. The focus of each embodiment is to describe the differences from other embodiments.

[0174] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0175] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0176] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute certain parts of the methods of various embodiments of this application.

[0177] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc.

[0178] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0179] The above description is merely a preferred embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to the embodiments described herein by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of protection of this specification.

Claims

1. An audio modulation method based on physiological signals, characterized in that, include: Obtain the current physiological signals of the target object; Based on the current physiological signal, physiological state features of multiple dimensions are extracted; Based on the physiological state characteristics of the multiple dimensions, the current state of the target object is determined; Based on the current state, at least one audio parameter of the audio to be played is dynamically adjusted to generate and output the audio to be played that matches the current state.

2. The audio modulation method based on physiological signals according to claim 1, characterized in that, The current physiological signal includes R-wave interval data, and the physiological state characteristics include at least two of the following: time-domain characteristics, frequency-domain characteristics, and nonlinear characteristics; The time-domain features include at least one of the following: statistical features of R-wave interval data, root mean square of the difference between adjacent R-wave intervals, proportion of adjacent R-wave interval differences greater than a preset threshold, and geometric features of the distribution of adjacent R-wave intervals. The frequency domain features include at least one of the following: the frequency band power of the frequency domain signal corresponding to the R-wave inter-interval data in multiple preset frequency bands, the total power of the frequency domain signal, the normalized power of the frequency domain signal, and the ratio of the power of at least two frequency bands. The nonlinear features include at least one of the following: sample entropy of R-wave interval data, self-similarity of R-wave interval data, geometric features of the Poincaré diagram corresponding to R-wave interval data, and complexity of R-wave interval data.

3. The audio modulation method based on physiological signals according to claim 2, characterized in that, The physiological state characteristics include at least nonlinear characteristics; Based on the current physiological signal, multiple dimensions of physiological state features are extracted, including: The R-wave interval data are matched based on a preset sample template to calculate the sample entropy of the R-wave interval data; Based on the R-wave interval data, the detrended fluctuation function corresponding to the R-wave interval data is calculated using the piecewise integration method, and the scaling exponent of the detrended fluctuation function is calculated as the self-similarity of the R-wave interval data. The Poincaré diagram is generated using an ellipse fitting method based on the R-wave interval data. The geometric features of the Poincaré diagram are calculated, and the geometric features include at least one of the following: the major axis, minor axis, and area of ​​the ellipse. Calculate the first complexity of the R-wave interval data at different time scales, the permutation entropy of the R-wave interval data, and the second complexity of the R-wave interval sequence corresponding to the R-wave interval data, and use the first complexity, the permutation entropy, and the second complexity as the complexity of the R-wave interval data.

4. The audio modulation method based on physiological signals according to claim 1, characterized in that, Based on the multiple dimensions of physiological state characteristics, the current state of the target object is determined, including: The feature weights of physiological state characteristics in each dimension are determined, and the current stress index of the target object is calculated as the current state based on the physiological state characteristics and feature weights of each dimension.

5. The audio modulation method based on physiological signals according to claim 4, characterized in that, Each dimension of physiological state characteristics includes the index values ​​corresponding to at least two physiological state indicators; Accordingly, the feature weights of physiological state characteristics in each dimension are determined. Based on the physiological state characteristics and feature weights in each dimension, the current stress index of the target object is calculated as the current state, including: The weights of each physiological state indicator are determined, and the current stress index is calculated as the current state based on the indicator values ​​and weights of each physiological state indicator.

6. The audio modulation method based on physiological signals according to claim 4 or 5, characterized in that, Also includes: Obtain the pressure index baseline corresponding to the target object. The pressure index baseline includes multiple preset pressure states and pressure index intervals corresponding to each pressure state. Based on the current pressure index and the pressure index baseline, the current pressure state of the target object is determined as the current state.

7. The audio modulation method based on physiological signals according to claim 6, characterized in that, The stress index baseline also includes the physiological state characteristic intervals corresponding to each stress state. The method further includes: Based on the pressure index baseline, determine the initial pressure state corresponding to the current pressure index; When the initial pressure state is determined to be in the boundary region of the pressure state category, a reference pressure state corresponding to the physiological state characteristic is determined based on the pressure index baseline. The current pressure state of the target object is determined as the current state based on the reference pressure state and the initial pressure state.

8. The audio modulation method based on physiological signals according to claim 6, characterized in that, Obtaining the baseline stress index corresponding to the target object includes: Select a portion of the historical physiological signals of the target object that meet a preset time threshold from the current time as the target physiological signal set; Calculate the historical stress index corresponding to each historical physiological signal in the target physiological signal set, and determine the distribution characteristics of the historical stress index; Based on the aforementioned distribution characteristics, the pressure index range corresponding to each pressure state is determined.

9. The audio modulation method based on physiological signals according to claim 1, characterized in that, Dynamically adjust at least one audio parameter of the audio to be played, including at least one of the following: Adjust the harmony of the audio to be played; Adjust the volume and / or frequency of each sound element in the audio to be played; Adjust the rhythm parameters of the audio to be played, wherein the rhythm parameters include at least one of the following: beat frequency and beat interval duration.

10. The audio modulation method based on physiological signals according to claim 9, characterized in that, Adjusting the harmonic harmony of the audio to be played includes: Determine the harmonic structure of the audio to be played, wherein the harmonic structure includes a chord sequence composed of multiple chord elements and the chord parameters of each chord element; Based on the mapping relationship between chords and scores, the harmonic weights corresponding to each chord, and the harmonic structure, the harmonic harmony of the audio to be played is determined. Determine the target harmony degree corresponding to the current state; Based on the difference between the stated harmonic harmony and the target harmony, a progressive conversion is used to adjust the harmonic structure of the audio to be played.

11. The audio modulation method based on physiological signals according to claim 10, characterized in that, Based on the mapping relationship between each harmony and the score, the harmonic weights corresponding to each harmony, and the harmonic structure, the harmonic harmony of the audio to be played is determined, including: Obtain the functional weight of each chord element in the chord sequence; Based on the chord sequence and the chord parameters of each chord element, calculate the context weight of each chord element; Based on the mapping relationship, the score corresponding to each chord element is determined; The harmonic harmony of the audio to be played is calculated based on the score corresponding to each chord element, the functional weight of each chord element, and the context weight of each chord element.

12. The audio modulation method based on physiological signals according to claim 9, characterized in that, Adjusting the volume and / or frequency of each sound element in the audio to be played, including: The instruments contained in the audio to be played are identified as the first sound elements; the instrument index corresponding to each instrument is obtained, the instrument index being used to characterize the degree of physiological influence of the corresponding instrument on the target object; based on the mapping relationship between stress state and instrument index, and in conjunction with preset volume adjustment rules, the volume of each instrument is adjusted; and / or, The spectrum of the audio to be played is determined as a second sound element. Based on the mapping relationship between the second sound element, pressure state, and frequency band gain value, the frequency band gain value of the audio to be played is adjusted; and / or, The spectral characteristics of each instrument in the audio to be played are determined as the third sound element. Based on the mapping relationship between the third sound element, the pressure state and the frequency band gain value of each instrument, the frequency band gain value of each instrument is adjusted.

13. The audio modulation method based on physiological signals according to claim 9, characterized in that, Adjust the rhythm parameters of the audio to be played, including: The multiple beat points contained in the audio to be played are determined, and the beat frequency of the audio to be played is determined based on the multiple beat points; Obtain the target beat frequency range corresponding to the current state; adjust the beat frequency based on the target beat frequency range, the beat frequency, and a preset beat frequency adjustment rule; and / or, Calculate the beat interval duration for each pair of beat points; determine the standard deviation and coefficient of variation for multiple beat points based on the beat interval duration; adjust at least one beat interval duration based on the difference between the beat frequency and the target beat frequency range; and / or, Obtain the beat heart rate synchronization rule corresponding to the current state, and adjust the beat frequency based on the beat heart rate synchronization rule.

14. An audio modulation device based on physiological signals, characterized in that, include: The acquisition module is used to acquire the current physiological signals of the target object; The extraction module is used to extract physiological state features in multiple dimensions based on the current physiological signal. The determination module is used to determine the current state of the target object based on the multiple dimensions of physiological state characteristics; An adjustment module is used to dynamically adjust at least one audio parameter of the audio to be played according to the current state, so as to generate and output the audio to be played that matches the current state.

15. An electronic device, characterized in that, include: A memory and a processor, the processor and the memory being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to implement the steps of the method according to any one of claims 1 to 13.

16. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 13.

Images