Sleep management method and device based on sound, equipment and medium

Abstract

The invention discloses a sound-based sleep management method and device, equipment and a medium. The method is applied to an intelligent sleep management system and can be applied to the fields of medical health or financial insurance and the like. The method comprises the following steps: collecting a sleep audio signal, and carrying out signal preprocessing on the sleep audio signal to obtain a target audio signal; performing time-frequency analysis and feature extraction on the target audio signal to obtain a sleep feature, wherein the sleep feature comprises a sleep stage; performing event detection on the sleep features through a preset neural network to obtain a sleep event; acquiring sleep quality scores of the sleep events and the sleep stages through a preset evaluation model, and constructing a sleep management scheme according to the sleep quality scores. By implementing the method provided by the invention, the problems that the sleep problem cannot be comprehensively considered, a sleep management scheme is difficult to generate and the sleep quality of a user is difficult to improve in the prior art can be solved, so that the chronic disease onset risk is reduced, the customer health level is improved, the insurance claim rate is indirectly reduced, and a win-win situation is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of smart homes and can be applied to fields such as medical health or finance and insurance. In particular, it relates to a sound-based sleep management method, device, equipment and medium. Background Art

[0002] In modern society, poor sleep quality has become a common phenomenon. The impact of sleep problems on life cannot be underestimated. Not only does it reduce the quality of daily life, but in the long term, it can also weaken immunity, increase the risk of chronic diseases, and pose numerous health risks. In the fields of pension and health insurance, integrating healthy sleep systems into products is of great significance, improving customer health and reducing claims, creating a win-win situation. However, current smart sleep detection systems on the market rely on wearable or contact devices, which can easily interfere with natural sleep and affect sleep quality. They only analyze body movement or breathing signals when detecting data sources, ignoring the correlation between sound and environmental noise, or they use a single sound analysis method that fails to comprehensively detect and correlate multiple information. Passive intervention capabilities are also limited. Traditional methods lack dynamic adjustment, or single environmental control methods lack multimodal coordination. These shortcomings make it difficult for existing technologies to comprehensively consider complex sleep issues, provide comprehensive health promotion solutions, and meet the needs of improving sleep and reducing health risks. Summary of the Invention

[0003] The embodiments of the present invention provide a sound-based sleep management method, apparatus, device, and medium, aiming to solve the problem that existing technologies cannot comprehensively consider complex sleep problems, are difficult to generate sleep management solutions, and help users improve their sleep quality.

[0004] In a first aspect, an embodiment of the present invention provides a sound-based sleep management method, which is applied to an intelligent sleep management system, comprising: collecting sleep audio signals according to a preset sampling frequency band, performing signal preprocessing on the sleep audio signals to obtain target audio signals, performing time-frequency analysis and feature extraction on the target audio signals to obtain sleep characteristics, wherein the sleep characteristics include sleep stages; performing event detection on the sleep characteristics through a preset neural network to obtain sleep events that occur during sleep; obtaining a sleep quality score by combining the sleep events and the sleep stages through a preset evaluation model, and constructing a corresponding sleep management plan based on the sleep quality score.

[0005] In a second aspect, an embodiment of the present invention further provides a sound-based sleep management device, which is applied to an intelligent sleep management system, and includes: an acquisition unit, for acquiring sleep audio signals according to a preset sampling frequency band, performing signal preprocessing on the sleep audio signals, and obtaining target audio signals; an extraction unit, for performing time-frequency analysis and feature extraction on the target audio signals to obtain sleep characteristics, wherein the sleep characteristics include sleep stages; a detection unit, for performing event detection on the sleep characteristics through a preset neural network, and obtaining sleep events that occur during the sleep process; a construction unit, for obtaining a sleep quality score by combining the sleep events and the sleep stages through a preset evaluation model, and constructing a corresponding sleep management plan based on the sleep quality score.

[0006] In a third aspect, an embodiment of the present invention further provides a computer device, which includes a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the above method when executing the computer program.

[0007] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the above method can be implemented.

[0008] Embodiments of the present invention provide a sound-based sleep management method, apparatus, device, and medium for use in intelligent sleep management systems. The method comprises: collecting sleep audio signals according to a preset sampling frequency band; performing signal preprocessing on the sleep audio signals to obtain a target audio signal; performing time-frequency analysis and feature extraction on the target audio signal to obtain sleep features, including sleep stages; applying event detection to the sleep features using a preset neural network to obtain sleep events that occurred during sleep; combining the sleep events with the sleep stages using a preset evaluation model to obtain a sleep quality score; and constructing a corresponding sleep management plan based on the sleep quality score. By collecting audio signals from a user during sleep and processing them to obtain sleep features and sleep stages, the present embodiment can monitor the user's sleep state in real time, facilitating event handling and sleep plan planning based on the sleep state. Sleep events occurring during sleep can be identified based on the sleep features, facilitating timely intervention and providing a deeper understanding of the user's sleep state. A sleep quality score is obtained based on the sleep events and sleep stages, thereby constructing a sleep management plan tailored to the user. This plan can comprehensively address complex sleep issues and improve sleep quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0010] Figure 1 A schematic diagram of a flow chart of a sound-based sleep management method provided in an embodiment of the present invention;

[0011] Figure 2 A schematic diagram of a sub-process of a sound-based sleep management method provided by an embodiment of the present invention;

[0012] Figure 3 A schematic diagram of a sub-process of a sound-based sleep management method provided by an embodiment of the present invention;

[0013] Figure 4 A schematic diagram of a sub-process of a sound-based sleep management method provided by an embodiment of the present invention;

[0014] Figure 5 A schematic diagram of a sub-process of a sound-based sleep management method provided by an embodiment of the present invention;

[0015] Figure 6 A schematic diagram of a sub-process of a sound-based sleep management method provided by an embodiment of the present invention;

[0016] Figure 7 A schematic diagram of a sub-process of a sound-based sleep management method provided by an embodiment of the present invention;

[0017] Figure 8 A schematic block diagram of a sound-based sleep management device provided by an embodiment of the present invention;

[0018] Figure 9 A schematic block diagram of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0020] It will be understood that when used in this specification and the appended claims, the terms “comprises” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0021] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the present invention. As used in the specification and appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0023] See also Figure 1 , Figure 1 A flow chart of a sound-based sleep management method provided in an embodiment of the present invention. The sound-based sleep management method in this embodiment can be applied to an intelligent sleep management system, wherein the system has a multi-microphone matrix and a piezoelectric sensor, and the piezoelectric sensor is applied to the bed to obtain the user's sleep state. The system is also connected to a smart pillow for sleep intervention. By adopting this method, a variety of human sleep information and sound characteristics can be collected, and the collected information can be analyzed and a dedicated management plan can be formulated to improve sleep quality and reduce the risk of illness during sleep, thereby improving sleep quality, reducing the risk of illness, and ensuring the health of users, which indirectly leads to a decrease in the claims ratio of pension insurance and health insurance, which is very beneficial to insurance companies, customers and the medical industry.

[0024] Figure 1 FIG. 1 is a flow chart of a sound-based sleep management method according to an embodiment of the present invention. As shown in the figure, the method includes the following steps S110-S140.

[0025] S110 : Collecting a sleep audio signal according to a preset sampling frequency band, performing signal preprocessing on the sleep audio signal, and obtaining a target audio signal.

[0026] In this embodiment, the sampling frequency band is the frequency range set when collecting audio signals. Sleep audio signals are collected according to the preset sampling frequency band. Specifically, the purpose of collecting sleep audio signals is to assess sleep quality, which can provide a basis for disease diagnosis and insurance risk assessment. Therefore, it is necessary to capture sound information closely related to sleep quality while eliminating irrelevant noise interference. In this embodiment, the preset sampling frequency band includes a low-frequency band and a high-frequency band. The low-frequency band (50-500Hz) focuses on capturing breathing and snoring, while the high-frequency band (1-2kHz) is used for coughing or environmental noise isolation. The audio sampling rate is ≥16kHz to collect accurate sleep audio signals. A multi-microphone array (such as a T-shaped or ring array) can be used to capture environmental sounds and breathing sounds. This is not limited to this method, as long as it can capture audio. The sleep audio signal is preprocessed. Specifically, the sleep audio signal can be preprocessed through filtering, noise reduction, and normalization, and the preprocessed audio signal is determined as the target audio signal. The target audio signal is obtained by sampling the sleep audio signal and performing preprocessing to remove noise and interference in the sleep audio signal, thereby improving the signal quality and signal-to-noise ratio, so as to more accurately extract feature information related to sleep quality.

[0027] In one embodiment, if Figure 2 As shown, the step S110 also includes steps S111-S113.

[0028] S111, performing noise suppression on the sleep audio signal using preset filtering parameters;

[0029] S112, dividing the noise-suppressed sleep audio signal according to a preset cutting time;

[0030] S113 , extracting features of the segmented sleep audio signal using a preset respiratory fundamental frequency and a preset classification algorithm to obtain the target audio signal.

[0031] In this embodiment, the preset filter parameters are designed based on the characteristics of sleep audio signals and common noise interference frequencies. The sleep audio signal is noise-suppressed using the preset filter parameters. Valid audio signals generated during sleep (such as breathing and snoring) are typically concentrated within a certain frequency range, while environmental noise (such as fan noise and traffic noise) may be distributed across different frequency bands. By setting the filter parameters, only the frequency range containing valid audio signals is allowed to pass, while noise components at other frequencies are filtered out. These parameters can be set based on commonly used parameters for normal sleep as reported in medical reports and are not limited to this. After noise suppression, the sleep audio signal is segmented according to a preset segmentation duration. Specifically, signal segmentation divides the continuous audio signal into multiple independent signal segments according to a specific time duration. The preset segmentation duration is determined based on analysis requirements and the characteristics of the audio signal. For example, the continuous audio signal can be segmented into multiple 30-second segments based on short-term energy detection (an audio signal processing method that segments a continuous audio signal into multiple short frames). The segmented sleep audio signal is subjected to feature extraction using a preset respiratory fundamental frequency and a preset classification algorithm to obtain the target audio signal. Specifically, the respiratory fundamental frequency refers to the basic frequency of the respiratory sound signal. Based on the frequency characteristics of the respiratory sound, a fundamental frequency range is set to extract respiratory-related features. The preset classification algorithm is blind source separation (ICA algorithm), which is an algorithm that recovers independent source signals from a mixed signal. It can extract the original independent source signal from the mixed signal. Audio filtering is performed using a preset fundamental frequency (0.1-0Hz), separating the respiratory waveform and extracting the respiratory signal in combination with blind source separation (ICA algorithm), thereby eliminating interference from body movement and obtaining an accurate target audio signal. By processing the sleep audio signal layer by layer to obtain the target audio signal, it is convenient to formulate a sleep management plan that suits the user's sleep state based on the target audio signal, and it can also help insurance companies provide customized health management plans. This can not only reduce customers' health risks, but also improve customer satisfaction and loyalty. It can also be combined with the medical system to provide patients with more accurate sleep disorder diagnosis and treatment plans.

[0032] S120 . Perform time-frequency analysis and feature extraction on the target audio signal to obtain sleep features, wherein the sleep features include sleep stages.

[0033] In this embodiment, the target audio signal is typically a sleep audio signal that has undergone preprocessing such as noise suppression and signal segmentation, and contains information related to sleep status, such as breathing sounds, snoring, and turning over. A time-frequency analysis is performed on the target audio signal to obtain its time-frequency representation (e.g., a spectrogram). A spectrogram can clearly display the energy distribution of the signal at different times and frequencies. For example, video analysis can be performed using methods such as short-time Fourier transform (STFT), wavelet transform (WT), and Hilbert-Huang transform (HHT) to obtain a spectrogram. Features related to sleep status are extracted from the spectrogram. Specifically, by detecting the frequency components corresponding to breathing sounds in the spectrogram, the respiratory rate is calculated. The sleep stage classification is performed based on the characteristic representations of different sleep stages (e.g., light sleep, deep sleep, and REM sleep) on the spectrogram, such as the intensity and duration of specific frequency components, thereby determining the user's sleep stage. If applied in a medical system, this can help doctors identify a patient's sleep characteristics, such as abnormal breathing rate and reduced heart rate variability, to assist in the diagnosis of sleep disorders (e.g., sleep apnea syndrome and insomnia). If applied to financial insurance systems, this data can help insurance companies assess customers' health risks based on extracted sleep characteristics and, combined with health data from medical reports, provide personalized health management advice and insurance product customization services. By performing time-frequency analysis and feature extraction on the target audio signal to obtain sleep characteristics, it facilitates comprehensive consideration of sleep issues based on these characteristics and the development of customized sleep management plans, ultimately improving the user's sleep quality.

[0034] In one embodiment, if Figure 3 As shown, the step S120 also includes steps S121-S124.

[0035] S121, generating a spectrogram by performing a preset Fourier transform on the target audio signal, and extracting an event feature signal according to the spectrogram;

[0036] S122, calculating the breathing interval difference of the target audio signal through a preset breathing algorithm to obtain a breathing signal event sequence;

[0037] S123, extracting respiratory variation features using slow wave features, and determining sleep stages by performing state classification based on the respiratory variation features;

[0038] S124: Determine the sleep characteristics according to the event characteristic signal, the respiratory signal event sequence, and the sleep stage.

[0039] In this embodiment, the Fourier transform is a method for converting a time-domain signal into a frequency-domain signal. A spectrogram can be generated by using the Fourier transform to display the energy distribution of the signal at different frequencies. The target audio signal is subjected to a preset Fourier transform to generate a spectrogram, and event characteristic signals are extracted from the spectrogram. Specifically, a short-time Fourier transform (STFT) is performed on the target audio signal to generate a time-frequency spectrum. By analyzing the frequency components and energy changes in the spectrogram, sleep-related event characteristic signals, such as breathing sounds, snoring, and turning over, are extracted from the spectrogram. For example, breathing characteristic signals are extracted in the low-frequency range (approximately 0.1-0.5 Hz), and snoring characteristic signals are extracted in the higher-frequency range. The target audio signal is subjected to a preset breathing algorithm to calculate respiratory interval differences and obtain a respiratory signal event sequence. Specifically, a breathing detection algorithm (such as peak detection or threshold detection) is used to identify respiratory events in the spectrogram, and the time difference between adjacent respiratory events is calculated to obtain a respiratory interval difference sequence. The respiratory interval differences are then smoothed and filtered to remove outliers, thereby generating a continuous respiratory signal event sequence. Respiratory variation features are extracted through slow wave features, and the sleep stages are determined by state division based on the respiratory variation features. Specifically, slow waves are usually associated with deep sleep stages. By extracting slow wave features in respiratory signals, the depth of sleep can be evaluated. The low-frequency components (slow waves) in the respiratory signal event sequence are analyzed, and their amplitude, duration and other features are extracted to determine the respiratory variation features. According to the respiratory variation features (such as slow wave features, respiratory interval changes), sleep is divided into different stages (such as light sleep, deep sleep, REM sleep). A classification algorithm (such as support vector machine, random forest, Balanced Bagging classifier) ​​or a rule system is used to divide the states according to the respiratory variation features to determine the different sleep stages during the sleep process. The sleep features are constructed according to the event feature signal, the respiratory signal event sequence and the sleep stage. By combining Fourier transform, respiratory algorithm and slow wave feature extraction technology, rich sleep features can be extracted from the sleep audio signal, and the user's sleep state can be deeply understood. Thereby improving the user's sleep quality.

[0040] S130: Perform event detection on the sleep characteristics through a preset neural network to obtain sleep events that occur during the sleep process.

[0041] In this embodiment, the preset neural network is a pre-selected neural network structure, such as a convolutional neural network (CNN), a recurrent neural network (RNN), a long short-term memory network (LSTM), or a Transformer model. The sleep characteristics are subjected to event detection through the preset neural network to obtain sleep events that occur during sleep. Specifically, the sleep characteristics are input into a trained neural network, and the neural network can perform event detection based on the sleep characteristics to obtain events that occur during sleep. For example, sleep events such as coughing, talking in sleep, and apnea can be detected. The sleep data of different patients collected in the medical system can be used as training data, and the neural network can be trained with the training data to identify and classify different sleep events. By performing event detection based on sleep characteristics, it is possible to understand what happens during the user's sleep, provide a basis for formulating a personalized sleep plan, and thus improve the user's sleep quality.

[0042] In one embodiment, if Figure 4 As shown, the step S130 also includes steps S131-S133.

[0043] S131, classifying and detecting the event characteristic signal using a preset convolutional neural network to determine whether a preset respiratory event occurs;

[0044] S132, detecting whether there is a respiratory signal interruption in the respiratory signal event sequence by a preset long-term and short-term neural network to determine whether a respiratory arrest event occurs;

[0045] S133: If the preset respiratory event and / or apnea event occurs, record it as the sleep event.

[0046] In this embodiment, the convolutional neural network is a deep learning model. The event characteristic signals are classified and detected by a preset convolutional neural network to determine whether a preset respiratory event has occurred. Specifically, the convolutional neural network is trained using a labeled data set to enable it to identify specific sleep events (such as snoring, turning over, etc.). The convolutional neural network detects and classifies the input event characteristic signals to determine whether they contain preset respiratory events such as snoring, coughing, and talking in sleep. The respiratory signal event sequence is detected by a preset long-term and short-term neural network to determine whether there is a respiratory signal interruption to determine whether an apnea event has occurred. Specifically, the long-term and short-term neural network is a recurrent neural network that can effectively process long-term dependencies in time series data. The long-term and short-term neural network is trained using a labeled apnea event data set to enable it to identify respiratory signal interruptions, thereby obtaining a preset long-term and short-term neural network. The generated respiratory signal event sequence is input into the preset long-term and short-term neural network to determine whether an apnea event has occurred, wherein the apnea event is an event such as apnea, respiratory interruption, etc. If any event occurs, it is recorded and classified as a sleep event. By combining the convolutional neural network and the long-term and short-term neural network, sleep events can be effectively detected and classified. It not only improves the accuracy and efficiency of sleep monitoring, helps improve users' sleep quality, but also provides important data support for medical diagnosis and health management.

[0047] In one embodiment, if Figure 5 As shown, step S130 further includes steps S1301-S1302.

[0048] S1301, determining whether the sleep event is a critical event;

[0049] S1302: If it is the critical event, perform sleep intervention according to the preset intervention method.

[0050] In this embodiment, the critical events refer to events that occur during sleep and may have a significant impact on individual health. In the insurance and medical fields, critical events generally include sleep interruptions caused by apnea, severe snoring, and frequent tossing and turning. For example, events such as apnea and prolonged snoring that occur during sleep are all critical events. Among them, key events can be set according to the type of event, frequency of event, and duration of event, which are not limited. It is determined whether there are critical events in the sleep events that occur during sleep. If a critical event occurs, such as an apnea event, sleep intervention is immediately performed according to the preset intervention method. Specifically, the intelligent sleep management system can control the connected smart pillow to vibrate to wake the user in time, or play the user's preset intervention music to intervene in the user's sleep. By performing sleep intervention in a timely manner when a critical event occurs, the user's safety is further protected while ensuring the user's sleep quality.

[0051] S140: Obtain a sleep quality score by combining the sleep events and the sleep stages through a preset evaluation model, and construct a corresponding sleep management plan according to the sleep quality score.

[0052] In this embodiment, the preset assessment model is constructed based on a large amount of sleep data and medical knowledge. It can comprehensively consider sleep events and sleep stages to calculate a sleep quality score. The sleep events and sleep stages are combined with the preset assessment model to obtain a sleep quality score. Specifically, the key factors related to sleep quality are determined in combination with medical knowledge, and the preset assessment model is constructed based on the key factors. In this embodiment, the key factors include sleep events and sleep stages. The sleep stages and the sleep events that occurred are weighted and calculated using the assessment model to obtain an overall sleep quality score. During the calculation, weights can be assigned to different dimensions (such as 30% deep sleep and 20% apnea frequency). A corresponding sleep management plan is constructed based on the sleep quality score. Specifically, a scoring system of 0-100 points can be set, for example: 80-100 points: excellent sleep quality; 60-79 points: good sleep quality but needs improvement; below 60 points: poor sleep quality, requiring special attention. A sleep management plan is created based on the different ranges of sleep quality scores. For example, if the sleep quality score is between 60 and 79, recommendations for adjusting sleep schedules and optimizing the sleep environment (such as playing white noise) are generated. By combining sleep events and sleep stage data, a sleep quality score is obtained through a preset evaluation model, and a corresponding sleep management plan is created based on the score, thereby improving individual sleep quality and health.

[0053] In one embodiment, if Figure 6 As shown, the step S140 also includes steps S141-S143.

[0054] S141, determining sleep efficiency and sleep duration according to the sleep stage;

[0055] S142, determining event frequency and respiratory stability according to the sleep duration and the sleep events;

[0056] S143: Determine the sleep quality score by using the preset evaluation model to measure the sleep efficiency, the event frequency, and the respiratory stability.

[0057] In this embodiment, the sleep efficiency refers to the ratio of actual sleep time to total bed time, which can be determined based on the ratio of actual sleep time to total bed time. Specifically. The actual sleep time (sleep duration) is determined based on the sum of the time of the sleep stages, and the total bed time is determined based on the information fed back by the bed sensor, thereby determining the sleep efficiency. The event frequency and respiratory stability are determined based on the sleep duration and the sleep events. Specifically, the event frequency refers to the number of sleep events that occur during the sleep time, such as apnea, awakening, turning over, etc. Respiratory stability refers to the frequency and severity of apnea or shallow breathing during the breathing process, which can be determined based on the number of apnea and hypopnea per hour in the sleep events. The sleep efficiency, the event frequency and the respiratory stability are used to determine the sleep quality score through the preset evaluation model, wherein the reference formula of the preset evaluation model is: Q = β0 + β1 × sleep efficiency + β2 × respiratory stability + β3 × event frequency + β4 ×

[0058] Environmental interference index, where weight parameters (β0, β1, β2, β3, β4) are derived through linear regression of accumulated data. By combining sleep stage, sleep duration, sleep events, and respiratory stability data, and using a preset quality assessment algorithm to calculate a sleep quality score, a more comprehensive assessment of the user's sleep quality can be achieved. This approach not only provides users with personalized sleep improvement recommendations but also provides important health assessment basis for the medical and insurance fields.

[0059] In one embodiment, if Figure 7 As shown, step S140 further includes steps S1401-S1402.

[0060] S1401: Before the user enters a sleep state again, the user's voice is recognized by a preset voice model to determine the user's sleep needs;

[0061] S1402: Search for corresponding sleep-aiding music in a preset music library according to the user's sleep needs and the constructed sleep management plan, and play the music.

[0062] In this embodiment, the preset voice model is a deep learning-based speech recognition model (e.g., an ASR model using an LSTM or Transformer architecture). Before the user enters a sleep state, it can interact with the user and receive user voice or other user commands. Based on the user's voice, the system determines the user's sleep needs. For example, the system uses natural language processing (NLP) within the preset voice model to analyze the intent of the user's voice, such as if the user desires a quieter environment. Based on the user's sleep needs and the established sleep management plan, the system searches for and plays corresponding sleep-inducing music in a preset music library. The preset music library is a database that categorizes and stores different music. A personalized sleep management plan library can be constructed based on the user's historical sleep data (e.g., sleep quality score, event frequency). If the user's voice request is quiet, the matching plan may include playing white noise. In this case, the system searches for and plays white noise in the music library. Music retrieval and playback can be performed using RAG (a technique that combines information retrieval and generation). By identifying the user's sleep needs through a preset voice model and combining personalized sleep management plans with a preset music library, accurate sleep-aid music recommendations and playback can be achieved, which can significantly enhance the user's sleep experience and thus improve the user's sleep quality.

[0063] Figure 8 FIG is a schematic block diagram of a sound-based sleep management device 200 provided by an embodiment of the present invention. Figure 8 As shown, corresponding to the above sound-based sleep management method, the present invention also provides a sound-based sleep management device. The sound-based sleep management device includes a unit for executing the above sound-based sleep management method. The device can be configured in a desktop computer, tablet computer, laptop computer, etc. For details, please refer to Figure 8 The sound-based sleep management device includes a collection unit 210 , an extraction unit 220 , a detection unit 230 and a construction unit 240 .

[0064] The collecting unit 210 is configured to collect a sleep audio signal according to a preset sampling frequency band, perform signal preprocessing on the sleep audio signal, and obtain a target audio signal.

[0065] In one embodiment, the acquisition unit 210 includes a suppression unit, a segmentation unit, and an acquisition unit.

[0066] a suppression unit, configured to suppress noise on the sleep audio signal through preset filtering parameters;

[0067] A segmentation unit, configured to segment the noise-suppressed sleep audio signal according to a preset segmentation time;

[0068] The acquisition unit is used to extract features of the segmented sleep audio signal through a preset respiratory fundamental frequency and a preset classification algorithm to obtain the target audio signal.

[0069] The extraction unit 220 is configured to perform time-frequency analysis and feature extraction on the target audio signal to obtain sleep features, wherein the sleep features include sleep stages.

[0070] In one embodiment, the extraction unit 220 includes a generation unit, a calculation unit, a division unit, and a determination unit.

[0071] a generating unit, configured to generate a spectrogram by performing a preset Fourier transform on the target audio signal, and extract an event feature signal according to the spectrogram;

[0072] a calculation unit, configured to calculate a breathing interval difference of the target audio signal through a preset breathing algorithm to obtain a breathing signal event sequence;

[0073] a classification unit, configured to extract respiratory variation features through slow wave features, and perform state classification to determine the sleep stage according to the respiratory variation features;

[0074] A determination unit is configured to determine the sleep characteristic according to the event characteristic signal, the respiratory signal event sequence, and the sleep stage.

[0075] The detection unit 230 is configured to perform event detection on the sleep characteristics through a preset neural network to obtain sleep events that occur during sleep.

[0076] In one embodiment, the detection unit 230 includes a first detection unit, a second detection unit, and a recording unit.

[0077] a first detection unit, configured to classify and detect the event characteristic signal through a preset convolutional neural network to determine whether a preset respiratory event occurs;

[0078] a second detection unit, configured to detect whether there is a respiratory signal interruption in the respiratory signal event sequence through a preset long-short term neural network, so as to determine whether a respiratory arrest event occurs;

[0079] The recording unit is configured to record the preset respiratory event and / or apnea event as the sleep event if it occurs.

[0080] In one embodiment, the detection unit 230 includes an event determination unit and an intervention unit.

[0081] an event determination unit, configured to determine whether the sleep event is a critical event;

[0082] The intervention unit is configured to perform sleep intervention according to a preset intervention method if the critical event occurs.

[0083] The construction unit 240 is configured to obtain a sleep quality score by combining the sleep events and the sleep stages through a preset evaluation model, and construct a corresponding sleep management plan according to the sleep quality score.

[0084] In one embodiment, the construction unit 240 includes a duration determination unit, a frequency determination unit, and a score determination unit.

[0085] a duration determination unit, configured to determine sleep efficiency and sleep duration according to the sleep stage;

[0086] a frequency determination unit, configured to determine an event frequency and a respiratory stability according to the sleep duration and the sleep event;

[0087] A score determination unit is configured to determine the sleep quality score by using the preset evaluation model to determine the sleep efficiency, the event frequency, and the respiratory stability.

[0088] It should be noted that those skilled in the art will clearly understand that the specific implementation process of the above-mentioned sound-based sleep management device 200 and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0089] The above-mentioned sound-based sleep management device can be implemented in the form of a computer program. The computer program can be used in Figure 9 Runs on the computer equipment shown.

[0090] See also Figure 9 , Figure 9 This is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device 500 can be a terminal or a server. The terminal can be a smart phone, tablet computer, laptop computer, desktop computer, personal digital assistant, wearable device, or other electronic device with communication capabilities. The server can be a standalone server or a server cluster consisting of multiple servers.

[0091] See Figure 9 The computer device 500 includes a processor 502 , a memory, and a network interface 505 connected via a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .

[0092] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions, which, when executed, may enable the processor 502 to perform a sound-based sleep management method.

[0093] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500.

[0094] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503 . When the computer program 5032 is executed by the processor 502 , the processor 502 can execute a sound-based sleep management method.

[0095] The network interface 505 is used to communicate with other devices through the network. Figure 9 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0096] The processor 502 is configured to run a computer program 5032 stored in the memory to implement the steps of the above method.

[0097] It should be understood that in the embodiment of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0098] Those skilled in the art will appreciate that all or part of the steps in the method of the above-described embodiment can be implemented by instructing the relevant hardware through a computer program. The computer program includes program instructions, which can be stored in a storage medium that is computer-readable. The program instructions are executed by at least one processor in the computer system to implement the steps in the method of the above-described embodiment.

[0099] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor performs the steps of the above method.

[0100] The storage medium may be any computer-readable storage medium that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk.

[0101] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0102] In the several embodiments provided herein, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the various units is merely a logical functional division, and actual implementation may employ other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be omitted or not implemented.

[0103] The steps in the methods of the embodiments of the present invention may be adjusted in order, combined, or deleted as needed. The units in the devices of the embodiments of the present invention may be combined, divided, or deleted as needed. Furthermore, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit.

[0104] If this integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the existing technology, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, terminal, or network device, etc.) to execute all or part of the steps of the method described in various embodiments of the present invention.

[0105] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and such modifications or substitutions are intended to be within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. A sound-based sleep management method, characterized in that: The method is applied to an intelligent sleep management system, and the method includes: Collecting a sleep audio signal according to a preset sampling frequency band, performing signal preprocessing on the sleep audio signal, and obtaining a target audio signal; Performing time-frequency analysis and feature extraction on the target audio signal to obtain sleep features, wherein the sleep features include sleep stages; Performing event detection on the sleep characteristics through a preset neural network to obtain sleep events that occur during sleep; The sleep events and the sleep stages are combined through a preset evaluation model to obtain a sleep quality score, and a corresponding sleep management plan is constructed according to the sleep quality score.

2. The method according to claim 1, characterized in that The step of performing signal preprocessing on the sleep audio signal to obtain a target audio signal includes: Performing noise suppression on the sleep audio signal through a preset filtering parameter; The noise-suppressed sleep audio signal is segmented according to the preset cutting time; The segmented sleep audio signal is subjected to feature extraction using a preset respiratory fundamental frequency and a preset classification algorithm to obtain the target audio signal.

3. The method according to claim 1, characterized in that The step of performing time-frequency analysis and feature extraction on the target audio signal to obtain sleep features includes: The target audio signal is subjected to a preset Fourier transform to generate a spectrogram, and an event feature signal is extracted according to the spectrogram; Calculate the breathing interval difference of the target audio signal through a preset breathing algorithm to obtain a breathing signal event sequence; Extracting respiratory variation features through slow wave features, and determining sleep stages by state classification based on the respiratory variation features; The sleep characteristic is determined according to the event characteristic signal, the respiratory signal event sequence, and the sleep stage.

4. The method according to claim 3, characterized in that The step of performing event detection on the sleep characteristics through a preset neural network includes: Classifying and detecting the event characteristic signal through a preset convolutional neural network to determine whether a preset respiratory event occurs; Detecting whether there is a respiratory signal interruption in the respiratory signal event sequence by a preset long-term and short-term neural network to determine whether a respiratory arrest event occurs; If the preset respiratory event and / or apnea event occurs, it is recorded as the sleep event.

5. The method according to claim 4, characterized in that After the step of recording the preset respiratory event and / or apnea event as the sleep event if it occurs, the method includes: Determining whether the sleep event is a critical event; If it is the critical event, sleep intervention is performed according to the preset intervention method.

6. The method according to claim 1, characterized in that The step of obtaining a sleep quality score by combining the sleep events and the sleep stages with a preset evaluation model includes: determining sleep efficiency and sleep duration based on the sleep stages; determining event frequency and respiratory stability according to the sleep duration and the sleep events; The sleep efficiency, the event frequency, and the respiratory stability are used to determine the sleep quality score through the preset evaluation model.

7. The method according to claim 1, characterized in that After the step of constructing a corresponding sleep management plan according to the sleep quality score, the method further includes: Before the user enters the sleep state again, the preset voice model is used to recognize the received user voice and determine the user's sleep needs; According to the user's sleep needs and the constructed sleep management plan, corresponding sleep-aiding music is searched and played in the preset music library.

8. A sound-based sleep management device, characterized in that: The device is applied to an intelligent sleep management system, and comprises: an acquisition unit, configured to acquire a sleep audio signal according to a preset sampling frequency band, perform signal preprocessing on the sleep audio signal, and obtain a target audio signal; an extraction unit, configured to perform time-frequency analysis and feature extraction on the target audio signal to obtain sleep features, wherein the sleep features include sleep stages; a detection unit, configured to perform event detection on the sleep characteristics through a preset neural network to obtain sleep events occurring during sleep; A construction unit is configured to obtain a sleep quality score by combining the sleep events and the sleep stages through a preset evaluation model, and to construct a corresponding sleep management plan according to the sleep quality score.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

10. A storage medium, characterized in that: The storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the method according to any one of claims 1 to 7 can be implemented.

Images