Snore detection method and system based on intelligent bed

Abstract

The embodiment of the invention provides a snoring detection method and system based on an intelligent bed, and the method comprises the steps: obtaining the sleep information of a user, collecting the sleep sound of the user based on the sleep information, and carrying out the snoring marking of the sleep sound; collecting sleep vibration signals of a user, preprocessing the sleep vibration signals to obtain signal segments, and extracting signal features; training samples are constructed through the sleep sound and the signal features, the signal features in the training samples are randomly combined, preset algorithms are selected to train the training samples under different random combinations, and in the training process, the preset algorithms and the random combinations are adjusted to obtain model training results corresponding to different preset algorithms and the random combinations; and selecting an exclusive training model of which the test precision is greater than a preset threshold in the model training result to be bound with the user, detecting a sleep vibration signal of the user based on the exclusive training model, and detecting whether snore exists or not.

Description

Technical Field

[0001] The present invention relates to the technical field of snoring signal recognition, and in particular to a snoring detection method and system based on a smart bed. Background Art

[0002] Snoring is a common sleep disorder that not only disrupts others' rest and affects social relationships, but can also be a sign of more serious health problems. To address this issue without invading the privacy of the bedroom, the smart bed is equipped with a piezoelectric sensor that can detect vibrations caused by snoring without disturbing the user. When the sensor detects snoring, the smart bed automatically and smoothly adjusts the angle of the head of the bed to help improve breathing, effectively reducing or eliminating snoring, and thus improving sleep quality.

[0003] However, although the snoring detection technology based on piezoelectric sensors has advantages in non-contact monitoring, it also has some limitations. First, it may be disturbed by the movement of the bed, resulting in false detection of snoring. In addition, due to technical limitations, it is impossible to accurately identify all types of snoring signals. In particular, for some special individuals, the signal characteristics generated when they snore may have obvious individual differences, resulting in the failure to detect such snoring signals. Summary of the invention

[0004] In view of the problems existing in the prior art, an embodiment of the present invention provides a snoring detection method and system based on a smart bed.

[0005] An embodiment of the present invention provides a snoring detection method based on a smart bed, the method comprising: Acquire the user's sleep information, collect the user's sleep sounds based on the sleep information, and mark the sleep sounds as snoring, wherein the sleep information includes the snoring level and the sleep time; Collecting a user's sleep vibration signal, preprocessing the sleep vibration signal to obtain a plurality of signal segments, and extracting signal features of the signal segments, wherein the signal features include: time domain features, frequency domain features, and time-frequency features; Construct training samples through sleep sounds and signal features after snoring marking, randomly combine the signal features in the training samples, select a preset algorithm to train the training samples under different random combinations, and during the training process, adjust the preset algorithm and random combination to obtain corresponding model training results under different preset algorithms and random combinations; An exclusive training model with a test accuracy greater than a preset threshold in the model training results is selected and bound to the user, and the user's sleep vibration signal is detected based on the exclusive training model to determine whether snoring exists.

[0006] In one embodiment, the method further comprises: When the number of users detected is greater than 1, the model corresponding to the preset algorithm is divided into models with different levels of sensitivity; Based on the snoring level selected by the user, a sensitivity model of the corresponding level is selected to detect the user's sleep vibration signal. When snoring is detected, a preset intervention action is triggered.

[0007] In one embodiment, the method further comprises: Collecting a vibration signal segment of the user every 20 seconds, and detecting whether the vibration signal segment contains the snoring mark; When the occurrence frequency of the snoring mark is greater than a preset frequency threshold, a preset intervention action is triggered.

[0008] In one embodiment, the method further comprises: Dividing the training samples into a training set and a test set; The selecting a preset algorithm to train the training samples under different random combinations includes: A preset algorithm is selected to train the training set under different random combinations, and the model performance is evaluated through the test set to obtain the test accuracy.

[0009] In one embodiment, the preprocessing comprises: Filtering, denoising, normalization and feature enhancement.

[0010] An embodiment of the present invention provides a snoring detection system based on a smart bed, the system comprising: An acquisition module, used to acquire the user's sleep information, collect the user's sleep sounds based on the sleep information, and mark the sleep sounds as snoring, wherein the sleep information includes the snoring level and the sleep time; An acquisition module is used to acquire a user's sleep vibration signal, pre-process the sleep vibration signal to obtain a plurality of signal segments, and extract signal features of the signal segments, wherein the signal features include: time domain features, frequency domain features, and time-frequency features; A training module is used to construct training samples through sleep sounds and signal features after snoring is marked, randomly combine the signal features in the training samples, select a preset algorithm to train the training samples under different random combinations, and during the training process, adjust the preset algorithm and random combination to obtain corresponding model training results under different preset algorithms and random combinations; The detection module is used to select an exclusive training model with a test accuracy greater than a preset threshold in the model training results and bind it to the user, and detect the user's sleep vibration signal based on the exclusive training model to determine whether snoring exists.

[0011] In one embodiment, the system further comprises: A division module, used for dividing the model corresponding to the preset algorithm into models of different sensitivity levels when the number of users detected is greater than 1; The selection module is used to select a sensitivity model of a corresponding level to detect the user's sleep vibration signal based on the snoring level selected by the user, and trigger a preset intervention action when snoring is detected.

[0012] In one embodiment, the system further comprises: A collection module, used for collecting a vibration signal segment of the user every 20 seconds, and detecting whether the vibration signal segment contains the snoring mark; The trigger module is used to trigger a preset intervention action when the occurrence frequency of the snoring mark is greater than a preset frequency threshold.

[0013] An embodiment of the present invention provides an electronic device, including a processor and a memory; The processor is connected to the memory; The memory is used to store executable program code; The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the method described in one or more embodiments.

[0014] An embodiment of the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the above-mentioned snoring detection method based on a smart bed are implemented.

[0015] In view of the above, in one or more embodiments of the present specification, the user's sleep information is obtained, the user's sleep sounds are collected based on the sleep information, and the sleep sounds are marked as snoring; the user's sleep vibration signal is collected, the sleep vibration signal is preprocessed to obtain multiple signal segments, and the signal features of the signal segments are extracted; the training samples are constructed through the sleep sounds and signal features after snoring marking, the signal features in the training samples are randomly combined, and the preset algorithm is selected to train the training samples under different random combinations. During the training process, the preset algorithm and the random combination are adjusted to obtain the corresponding model training results under different preset algorithms and random combinations; the exclusive training model with a test accuracy greater than the preset threshold in the model training results is selected to bind to the user, and the user's sleep vibration signal is detected based on the exclusive training model to determine whether snoring exists. In this way, the specific characteristics of the individual can be used to create an exclusive individual model through separate training to achieve more accurate snoring detection, and personalized intervention measures can be configured to ensure that each user can obtain the snoring solution that best suits him or her. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, 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 creative work.

[0017] Figure 1 This is a flow chart of a snoring detection method based on a smart bed provided in one embodiment of this specification.

[0018] Figure 2 It is a schematic diagram of a model training provided by an embodiment of this specification.

[0019] Figure 3 This is a flow chart of a snoring detection and intervention process provided by an embodiment of this specification.

[0020] Figure 4 This is a flow chart of another snoring detection method based on a smart bed provided in one embodiment of this specification.

[0021] Figure 5 It is a structural diagram of a snoring detection system based on a smart bed provided in one embodiment of this specification.

[0022] Figure 6 It is a structural schematic diagram of an electronic device provided by an embodiment of this specification. DETAILED DESCRIPTION

[0023] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and is not a limitation of the scope of protection, applicability or examples set forth in the claims. The functions and arrangements of the elements discussed can be changed without departing from the scope of protection of the contents of this specification. Various examples can omit, replace or add various processes or components as needed. For example, the described method can be performed in an order different from the described order, and various steps can be added, omitted or combined. In addition, the features described relative to some examples can also be combined in other examples.

[0024] As used herein, the term "including" and its variations represent open terms, meaning "including but not limited to". The term "based on" means "based at least in part on". The terms "one embodiment" and "an embodiment" mean "at least one embodiment". The term "another embodiment" means "at least one other embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other definitions may be included below, whether explicit or implicit. Unless the context clearly indicates otherwise, the definition of a term is consistent throughout the specification.

[0025] like Figure 1 As shown, an embodiment of the present invention provides a snoring detection method based on a smart bed, comprising: Step S102, obtaining the user's sleep information, collecting the user's sleep sounds based on the sleep information, and marking the sleep sounds as snoring, wherein the sleep information includes the snoring level and the sleep time.

[0026] Specifically, in order to intelligently detect or intervene in the user's snoring information, the user's sleep information can be obtained to provide a data basis for subsequent steps. The sleep information can be obtained by active acquisition or passive input. Active acquisition can be to obtain the user's sleep audio through the recording function of the bound terminal after connecting to the user's bound terminal. Passive input can be for the user to directly input his or her snoring situation (such as none, mild, moderate or severe), as well as sleep time, etc. through the bound terminal. Obtaining sleep information can further determine the intervention time of the smart bed on the user's snoring behavior.

[0027] Furthermore, after determining the user's sleep information, the user's sleep sounds can be collected based on the sleep information, such as recording the user's entire sleep process, or dividing the recording into segments of 20 seconds each, which helps to reduce the amount of data in a single processing unit. Each segment is then analyzed using an audio classification model. The audio model can be a model that has been trained based on snoring samples, and can directly determine whether the user is snoring based on the sound in the audio segment, and after making a snoring judgment, the sleep sound is marked with a snoring mark, such as a mark of "1" for snoring and "0" for no snoring.

[0028] In addition, when marking sleep sounds as snoring, the start time of each audio clip should also be recorded. Recording the start time is helpful in determining when the user's snoring occurs and can help understand the user's snoring pattern, such as whether snoring is more frequent in a specific sleep stage.

[0029] Step S104, collecting the user's sleep vibration signal, preprocessing the sleep vibration signal to obtain multiple signal segments, and extracting signal features of the signal segments, the signal features including: time domain features, frequency domain features and time-frequency features.

[0030] Specifically, the non-contact pressure sensor on the smart bed can be used to collect vibration signals generated by the user when sleeping. The sensor can sense subtle changes in the mattress, such as the user's breathing pattern, turning over, and so on. In addition, in order to ensure the time alignment of the vibration signal and the audio clip, the timestamps of the two need to be accurately synchronized. For comparing audio clips, the vibration signal can also be sliced ​​into segments every 20 seconds, and the start time of these segments is consistent with the start time of the previously uploaded recording mark segment, so as to achieve synchronous analysis of the two.

[0031] Furthermore, the sleep vibration signal is preprocessed, including filtering, denoising, normalization, and feature enhancement. Filtering can remove noise of a specific frequency through a digital filter (such as a band-stop filter) to improve signal quality, and denoising can use a digital filter to remove environmental noise and other interfering signals. Normalization unifies the data. Feature enhancement includes smoothing, trend removal, and so on.

[0032] Furthermore, the signal features of the preprocessed multiple signal segments are extracted, wherein the signal features include time domain features, which extract statistical characteristics from the original vibration signal, such as mean, standard deviation, maximum value, minimum value, peak-to-peak value, zero-crossing rate, etc. Time domain features can identify regularities and abnormalities in signals, such as periodic vibrations caused by snoring. Frequency domain features, through Fourier transform (FFT) or other spectrum analysis methods, convert time domain signals to frequency domain, and then calculate power spectral density (PSD), main frequency, bandwidth and other features. Frequency domain features can play a greater role in distinguishing different types of vibration sources (such as snoring and other activities). Time-frequency features, combining time domain and frequency domain information, use short-time Fourier transform (STFT), wavelet transform and other technologies to generate time-frequency representation. The time-frequency feature extraction method is suitable for capturing the frequency components of the signal that change over time, which helps to more accurately identify snoring events and their development process.

[0033] Step S106, constructing training samples through sleep sounds and signal features after snoring marking, randomly combining the signal features in the training samples, selecting a preset algorithm to train the training samples under different random combinations, and during the training process, adjusting the preset algorithm and random combination to obtain corresponding model training results under different preset algorithms and random combinations.

[0034] Specifically, training samples are constructed based on sleep sounds and signal features after snoring is marked, including collecting all segment labels marked as snoring (1) or non-snoring (0). The segment labels can be used to train and evaluate machine learning models during model training. Then all time domain, frequency domain and time-frequency features are extracted from the vibration signal of the smart bed, including but not limited to time domain features such as mean, standard deviation, maximum value, minimum value, zero-crossing rate; frequency domain features such as power spectral density, main frequency, bandwidth; and time-frequency features obtained by short-time Fourier transform or wavelet transform. These features are combined with the audio segment labels of the corresponding time into new training samples.

[0035] In addition, the training samples can be divided into training sets and test sets. The training set is used to train the model, while the test set is used to evaluate the model performance.

[0036] Furthermore, the signal features in the training samples are randomly combined, for example, according to the data domain, combination 1: mean, maximum value, minimum value, zero-crossing rate (time domain features); combination 2: power spectrum density, main frequency, bandwidth, frequency center (frequency domain features); combination 3: the first five coefficients in the short-time Fourier transform result (time-frequency features); combination 4: combination 1 + combination 2 (time domain + frequency domain). After determining the random combination, the training samples under different random combinations are trained according to the preset algorithm. The preset algorithm can be an algorithm library, for example, the algorithm library includes algorithms in multiple directions, such as XGBoost, linear regression, SVM (support vector machine), logistic regression, decision tree, random forest, naive Bayes, etc. Each algorithm has its own characteristics and applicable scenarios. For example, random forest is suitable for processing high-dimensional data and is not easy to overfit; SVM works better for data with small sample sizes; XGBoost is good at processing large-scale data sets. The process of adjusting the preset algorithm and random combination for random model training can be, for example, taking random forest as an example, combining the features of combination 1, training the training set, and using the test set to evaluate the model performance to obtain the initial test accuracy Acc1. Then adjust the preset algorithm and random combination, traverse all possible combinations, and use other algorithms to combine and pair with other features, such as using other algorithms to pair the remaining three sets of features. After each iteration, a new test accuracy value will be obtained and recorded to form a test accuracy sequence Acc1, Acc2, ..., Accmxn, where m is the number of algorithms and n is the number of feature combinations, such as Figure 2 shown.

[0037] Step S108, selecting an exclusive training model with a test accuracy greater than a preset threshold in the model training results and binding it to the user, and detecting the user's sleep vibration signal based on the exclusive training model to determine whether snoring exists.

[0038] Specifically, after determining the test accuracy values ​​of all model training results, an exclusive training model with a test accuracy greater than a preset threshold can be selected and bound to the user, such as the most accurate batch of training models or a training model, and determined as the user's exclusive model, which becomes the basis for the user's personalized snoring detection. The smart bed can provide users with specific snoring detection through this model, including detecting the user's sleep vibration signal, specifically detecting the user's sleep vibration signal. When the user's sleep signal corresponds to a label of 1 (snoring), it means that the user is snoring, and then triggering the corresponding intervention action, such as the smart bed will automatically perform the head of the bed lifting action to help improve breathing and reduce or stop snoring. It can also be set by the user, such as generating slight vibrations, or making small sounds, etc., and there are no restrictions here.

[0039] Furthermore, the process of detecting and intervening whether the user is snoring can be as follows: Figure 3 As shown, based on the user-specific training model, a vibration signal segment can be analyzed every 20 seconds. If snoring is detected, a snoring mark is output. If the frequency of the snoring mark is greater than the preset frequency threshold, such as 3 snoring marks appearing in 1 minute, it means that the user is snoring. Then trigger the corresponding personalized intervention action, such as raising the head of the bed. This high-frequency detection can ensure timely response to snoring events while maintaining a low false alarm rate.

[0040] In addition, when the number of users detected is greater than 1, according to the characteristics of the group snoring signal, the general detection model trained can be divided into low-sensitivity model, medium-sensitivity model and high-sensitivity model according to different parameters. When the model is trained, the positive and negative samples are balanced. For the low-sensitivity model, the precision rate of snoring detection is prioritized during model training, and the recall rate is secondary; the medium-sensitivity model is a balance between the precision rate and the recall rate; the high-sensitivity model prioritizes the recall rate during model training, and the precision rate is secondary. The three models with different sensitivities can be the same algorithm model but with different model parameters, such as a decision tree model, but the tree depth, number of leaf nodes and other parameters are different; or they can be three different algorithm models, such as using a support vector machine model, a decision tree model and a Bayesian model respectively. The choice of model is ultimately determined based on the user's understanding of his or her own situation. If the user selects mild snoring on the APP, the model automatically selects low sensitivity, which can effectively reduce the probability of false snoring detection; if the user selects moderate snoring, the model automatically selects medium sensitivity to balance the results of snoring detection; if the user selects severe snoring, the model automatically selects high sensitivity to ensure that the user can be effectively intervened when snoring. When snoring is detected, the preset intervention action can be triggered, such as the smart bed automatically performing the head of the bed lifting action to help improve breathing and reduce or stop snoring. This achieves personalized snoring detection and intervention for multiple people.

[0041] Specifically, in this embodiment, the snoring detection method based on the smart bed can be as follows: Figure 4 As shown, it is composed of 7 units, including a user basic information acquisition unit and a recording marking unit in step S102, which acquire the user's sleep information, collect the user's sleep sound based on the sleep information, and mark the snoring sound through the sleep sound. And the smart bed signal acquisition unit and data processing unit in step S104, collect the user's sleep vibration signal, pre-process the sleep vibration signal, obtain multiple signal segments, and extract the signal features of the signal segments. And a snoring model training unit, which constructs training samples through sleep sounds and signal features, randomly combines the signal features in the training samples, selects a preset algorithm to train the training samples under different random combinations, and adjusts the preset algorithm and random combination during the training process to obtain the corresponding model training results under different preset algorithms and random combinations. Finally, it includes an individual snoring detection unit and a snoring intervention unit, which selects a dedicated training model with a test accuracy greater than a preset threshold in the model training result and binds it to the user, and detects the user's sleep vibration signal based on the dedicated training model. When snoring is detected, a preset intervention action is triggered.

[0042] The embodiment of the present invention provides a snoring detection method based on a smart bed, which obtains the user's sleep information, collects the user's sleep sound based on the sleep information, and marks the snoring through the sleep sound; collects the user's sleep vibration signal, pre-processes the sleep vibration signal, obtains multiple signal segments, and extracts the signal features of the signal segments; constructs training samples through sleep sounds and signal features, randomly combines the signal features in the training samples, selects a preset algorithm to train the training samples under different random combinations, and during the training process, adjusts the preset algorithm and random combination to obtain the corresponding model training results under different preset algorithms and random combinations; selects an exclusive training model with a test accuracy greater than a preset threshold in the model training result and binds it to the user, and detects the user's sleep vibration signal based on the exclusive training model, and triggers a preset intervention action when snoring is detected. In this way, the specific characteristics of the individual can be used to create an exclusive individual model through separate training to achieve more accurate snoring detection, and personalized intervention measures can be configured to ensure that each user can obtain the snoring solution that best suits him or her.

[0043] See also Figure 5 , Figure 5 Schematic diagram of a snoring detection system based on a smart bed provided in an embodiment of the present application. Figure 5 As shown, the system comprises: The acquisition module S502 is used to acquire the user's sleep information, collect the user's sleep sounds based on the sleep information, and mark the sleep sounds as snoring, wherein the sleep information includes the snoring level and the sleep time; The acquisition module S504 is used to acquire the user's sleep vibration signal, pre-process the sleep vibration signal to obtain multiple signal segments, and extract signal features of the signal segments, wherein the signal features include: time domain features, frequency domain features, and time-frequency features; The training module S506 is used to construct training samples based on the sleep sounds and signal features after snoring is marked, randomly combine the signal features in the training samples, select a preset algorithm to train the training samples under different random combinations, and during the training process, adjust the preset algorithm and the random combination to obtain the corresponding model training results under different preset algorithms and random combinations; The detection module S508 is used to select an exclusive training model with a test accuracy greater than a preset threshold in the model training results and bind it to the user, and detect the user's sleep vibration signal based on the exclusive training model to determine whether there is snoring.

[0044] In another embodiment, a snoring detection system based on a smart bed further includes: A division module, used for dividing the model corresponding to the preset algorithm into models of different sensitivity levels when the number of users detected is greater than 1; The selection module is used to select a sensitivity model of a corresponding level to detect the user's sleep vibration signal based on the snoring level selected by the user, and trigger a preset intervention action when snoring is detected.

[0045] In another embodiment, a snoring detection system based on a smart bed further includes: A collection module, used for collecting a vibration signal segment of the user every 20 seconds, and detecting whether the vibration signal segment contains the snoring mark; The trigger module is used to trigger a preset intervention action when the occurrence frequency of the snoring mark is greater than a preset frequency threshold.

[0046] Those skilled in the art can clearly understand that the technical solutions of the embodiments of the present application can be implemented with the help of software and / or hardware. The "unit" and "module" in this specification refer to software and / or hardware that can independently complete or cooperate with other components to complete specific functions, where the hardware can be, for example, a field programmable gate array (FPGA), an integrated circuit (IC), etc.

[0047] Each processing unit and / or module of the embodiments of the present application may be implemented by an analog circuit that implements the functions described in the embodiments of the present application, or may be implemented by software that executes the functions described in the embodiments of the present application.

[0048] See also Figure 6 , which shows a schematic diagram of the structure of an electronic device involved in an embodiment of the present application, the electronic device can be used to implement Figure 1 The method in the embodiment shown. Figure 6 As shown, the electronic device 600 may include: at least one processor 601 , at least one network interface 604 , a user interface 603 , a memory 605 , and at least one communication bus 602 .

[0049] The communication bus 602 is used to realize the connection and communication between these components.

[0050] The user interface 603 may include a display screen (Display) and a camera (Camera). The optional user interface 603 may also include a standard wired interface and a wireless interface.

[0051] The network interface 604 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0052] Among them, the processor 601 may include one or more processing cores. The processor 601 uses various interfaces and lines to connect various parts of the entire electronic device 600, and executes various functions and processes data of the electronic device 600 by running or executing instructions, programs, code sets or instruction sets stored in the memory 605, and calling data stored in the memory 605. Optionally, the processor 601 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 601 can integrate one or a combination of a processor (Central Processing Unit, CPU), an image processor (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 601, and it can be implemented separately through a chip.

[0053] Among them, the memory 605 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 605 includes a non-transitory computer-readable storage medium. The memory 605 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 605 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 605 may also be optionally at least one storage device located away from the aforementioned processor 601. As Figure 6 As shown, the memory 605 as a computer storage medium may include an operating system, a network communication module, a user interface module, and program instructions.

[0054] exist Figure 6In the electronic device 600 shown, the user interface 603 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 601 can be used to call the interactive application based on image generation stored in the memory 605, and specifically perform the following operations: obtain the user's sleep information, collect the user's sleep sound based on the sleep information, and mark the snoring through the sleep sound; collect the user's sleep vibration signal, pre-process the sleep vibration signal to obtain multiple signal segments, and extract the signal features of the signal segments; construct training samples through sleep sounds and signal features, randomly combine the signal features in the training samples, select a preset algorithm to train the training samples under different random combinations, and during the training process, adjust the preset algorithm and random combination to obtain corresponding model training results under different preset algorithms and random combinations; select an exclusive training model with a test accuracy greater than a preset threshold in the model training result to bind to the user, and detect the user's sleep vibration signal based on the exclusive training model, and when snoring is detected, trigger a preset intervention action.

[0055] The present application also provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the steps of the above method are implemented. The computer-readable storage medium may include, but is not limited to, any type of disk, including a floppy disk, an optical disk, a DVD, a CD-ROM, a micro drive, and a magneto-optical disk, a ROM, a RAM, an EPROM, an EEPROM, a DRAM, a VRAM, a flash memory device, a magnetic card or an optical card, a nanosystem (including a molecular memory IC), or any type of medium or device suitable for storing instructions and / or data.

[0056] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0057] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0058] In the several embodiments provided in the present application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of the units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0059] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0060] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0061] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned memory includes: U disk, read-only memory (ROM), random access memory (RAM), mobile hard disk, disk or optical disk and other media that can store program codes.

[0062] A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by entering a program to instruct related hardware, and the program may be stored in a computer-readable memory, which may include a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0063] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A snoring detection method based on a smart bed, the method comprising: Acquire the user's sleep information, collect the user's sleep sounds based on the sleep information, and mark the sleep sounds as snoring, wherein the sleep information includes the snoring level and the sleep time; Collecting a user's sleep vibration signal, preprocessing the sleep vibration signal to obtain a plurality of signal segments, and extracting signal features of the signal segments, wherein the signal features include: time domain features, frequency domain features, and time-frequency features; Construct training samples through sleep sounds and signal features after snoring marking, randomly combine the signal features in the training samples, select a preset algorithm to train the training samples under different random combinations, and during the training process, adjust the preset algorithm and random combination to obtain corresponding model training results under different preset algorithms and random combinations; An exclusive training model with a test accuracy greater than a preset threshold in the model training results is selected and bound to the user, and the user's sleep vibration signal is detected based on the exclusive training model to determine whether snoring exists.

2. The method according to claim 1, characterized in that The method further comprises: When the number of users detected is greater than 1, the model corresponding to the preset algorithm is divided into models with different levels of sensitivity; Based on the snoring level selected by the user, a sensitivity model of the corresponding level is selected to detect the user's sleep vibration signal. When snoring is detected, a preset intervention action is triggered.

3. The method according to claim 1, characterized in that The method further comprises: Collecting a vibration signal segment of the user every 20 seconds, and detecting whether the vibration signal segment contains the snoring mark; When the occurrence frequency of the snoring mark is greater than a preset frequency threshold, a preset intervention action is triggered.

4. The method according to claim 1, characterized in that The method further comprises: Dividing the training samples into a training set and a test set; The selecting a preset algorithm to train the training samples under different random combinations includes: A preset algorithm is selected to train the training sets under different random combinations, and the model performance is evaluated through the test set to obtain the test accuracy.

5. The method according to claim 1, characterized in that The preprocessing comprises: Filtering, denoising, normalization and feature enhancement.

6. A snoring detection system based on a smart bed, characterized in that: The system comprises: An acquisition module, used to acquire the user's sleep information, collect the user's sleep sounds based on the sleep information, and mark the sleep sounds as snoring, wherein the sleep information includes the snoring level and the sleep time; An acquisition module is used to acquire a user's sleep vibration signal, pre-process the sleep vibration signal to obtain a plurality of signal segments, and extract signal features of the signal segments, wherein the signal features include: time domain features, frequency domain features, and time-frequency features; A training module is used to construct training samples through sleep sounds and signal features after snoring is marked, randomly combine the signal features in the training samples, select a preset algorithm to train the training samples under different random combinations, and during the training process, adjust the preset algorithm and random combination to obtain corresponding model training results under different preset algorithms and random combinations; The detection module is used to select an exclusive training model with a test accuracy greater than a preset threshold in the model training results and bind it to the user, and detect the user's sleep vibration signal based on the exclusive training model to determine whether snoring exists.

7. The system according to claim 6, characterized in that The system further comprises: A division module, used for dividing the model corresponding to the preset algorithm into models of different sensitivity levels when the number of users detected is greater than 1; The selection module is used to select a sensitivity model of a corresponding level to detect the user's sleep vibration signal based on the snoring level selected by the user, and trigger a preset intervention action when snoring is detected.

8. The system according to claim 6, characterized in that The system further comprises: A collection module, used for collecting a vibration signal segment of the user every 20 seconds, and detecting whether the vibration signal segment contains the snoring mark; The trigger module is used to trigger a preset intervention action when the occurrence frequency of the snoring mark is greater than a preset frequency threshold.

9. An electronic device comprising a processor and a memory; The processor is connected to the memory; The memory is used to store executable program code; The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the method according to any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method according to any one of claims 1 to 5.

Images