Method and apparatus for determining sleep posture of user and providing sleep posture information using deep learning model

Abstract

The present invention relates to a method and an apparatus for determining a sleep posture of a user and providing sleep posture information using a deep learning model, the method performed by a sleep monitoring device comprising the steps of: receiving sensor data from a sleep sensor; inputting the sensor data into a pre-trained deep learning model; determining a sleep posture of a user on the basis of output data outputted from the pre-trained deep learning model; measuring a sleep time and a non-sleep time according to the sleep posture; and providing the sleep posture information to the user terminal.

Description

Method and device for determining a user's sleeping posture and providing sleeping posture information using a deep learning model

[0001] The present invention relates to a method and a device for determining a user's sleeping posture and providing sleeping posture information using a deep learning model, and more particularly, to a method and a device for providing sleeping posture that can determine a user's sleeping posture using a pressure sensor included in a sleep sensor, monitor the user's sleeping posture in real time, and provide information about the user's sleeping posture.

[0002] Research on recognizing various human activities is being conducted in a variety of fields, including computer vision and pattern recognition. Behavior recognition estimates user behavior and transmits or requests information, utilizing various sensors to predict behavior.

[0003] With the growing interest in wellness, research is actively underway to identify repetitive behaviors in daily life, such as sleeping and sitting posture. Sleep, which accounts for more than a third of the day, not only serves to relieve fatigue, but also provides the only restorative time for the vertebrae, discs, muscles, and ligaments that support the body throughout the day. Improper sleeping posture is a major cause of spinal disorders, muscle pain, and sleep apnea, so maintaining good sleeping habits is crucial. Consequently, research into sleep posture analysis is on the rise.

[0004] Therefore, there is a need for further discussion on how to utilize deep learning technology to provide information about sleep posture more effectively and efficiently for sleep posture analysis.

[0005] The background technology described above is technical information that the inventor possessed for the purpose of deriving the present invention or acquired during the process of deriving the present invention, and cannot necessarily be said to be technology known to the general public prior to the filing of the present invention.

[0006] The problem to be solved by the disclosure of the present invention is to determine a user's sleeping posture using only a pressure sensor and provide the user with sleeping posture information related thereto.

[0007] In addition, the problem to be solved through the disclosure of the present invention is to provide a deep learning technology for more effective and efficient sleep posture determination by continuously learning a deep learning model.

[0008] A method performed by a sleep monitoring device for a problem to be solved through some embodiments of the present invention may include a step of receiving sensor data from a sleep sensor, a step of inputting the sensor data into a pre-trained deep learning model, a step of determining a user's sleeping posture based on output data output from the pre-trained deep learning model, a step of measuring a sleeping time and a non-sleeping time according to the sleeping posture, and a step of providing sleeping posture information to a user terminal.

[0009] In one embodiment, the step of providing sleep posture information to the user terminal may further include the step of receiving feedback information regarding the sleep posture from the user terminal.

[0010] In one embodiment, the step of providing sleep posture information to the user terminal may further include the step of setting deep learning model learning weights for feedback information.

[0011] In one embodiment, the step of providing sleep posture information to the user terminal may further include a step of learning a deep learning model according to learning weights.

[0012] In one embodiment, the deep learning model may be characterized by including at least one convolutional layer, a BNResidual Blocks layer, and an SPP layer.

[0013] In one embodiment, the BNResidual Blocks layer in the deep learning model may be composed of at least one convolutional layer.

[0014] In one embodiment, the SPP layer in the deep learning model may be configured to include at least one convolutional layer and at least one max pooling layer.

[0015] According to the problem solving means of the present invention described above, the user's sleeping posture can be determined using only a pressure sensor, and sleeping posture information related thereto can be provided to the user.

[0016] In addition, according to the problem solving means of the present invention, a deep learning model can be continuously trained to provide a deep learning technology for more effective and efficient sleep posture determination.

[0017] FIG. 1 illustrates an exemplary environment in which a sleep monitoring device according to some embodiments of the present disclosure may be applied.

[0018] FIG. 2 is a flowchart illustrating an operation of a sleep monitoring device for determining a user's sleeping posture according to some embodiments of the present disclosure.

[0019] FIG. 3 is a flowchart for a specific description of steps for providing sleep posture information to a user terminal according to some embodiments of the present disclosure.

[0020] FIG. 4 is an exemplary diagram of a deep learning model architecture of the present invention according to some embodiments of the present disclosure.

[0021] FIG. 5 is a diagram of an exemplary computing device that may implement a device and / or system according to various embodiments of the present disclosure.

[0022] Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the attached drawings. The advantages and features of the present disclosure, and methods for achieving them, will become clearer with reference to the embodiments described in detail below together with the attached drawings. However, the technical idea of ​​the present disclosure is not limited to the following embodiments and may be implemented in various different forms. The following embodiments are provided only to complete the technical idea of ​​the present disclosure and to fully inform those skilled in the art of the present disclosure of the scope of the present disclosure, and the technical idea of ​​the present disclosure is defined only by the scope of the claims.

[0023] When assigning reference numerals to components in each drawing, it should be noted that identical components are assigned the same numerals whenever possible, even if they appear on different drawings. Furthermore, when describing the present disclosure, if a detailed description of a related known configuration or function is deemed likely to obscure the gist of the present disclosure, such detailed description will be omitted.

[0024] Unless otherwise defined, all terms (including technical and scientific terms) used herein may be used in a meaning that can be commonly understood by a person of ordinary skill in the art to which this disclosure belongs. In addition, terms defined in commonly used dictionaries shall not be interpreted ideally or excessively unless explicitly specifically defined. The terminology used herein is for the purpose of describing embodiments and is not intended to limit the present disclosure. In this specification, the singular also includes the plural unless specifically stated otherwise in the phrase.

[0025] Additionally, terms such as first, second, A, B, (a), (b), etc. may be used to describe components of the present disclosure. These terms are only intended to distinguish the components from other components, and the nature, order, or sequence of the components are not limited by the terms. When it is described that a component is "connected," "coupled," or "connected" to another component, it should be understood that the component may be directly connected or connected to the other component, but another component may also be "connected," "coupled," or "connected" between each component.

[0026] The terms "comprises" and / or "comprising" as used in the specification do not exclude the presence or addition of one or more other components, steps, operations and / or elements.

[0027] Hereinafter, various embodiments of the present disclosure will be described in detail with reference to the attached drawings.

[0028] In addition, when describing the components of the present invention, terms such as first, second, A, B, (a), (b), etc. may be used. These terms are only to distinguish the components from other components, and the nature, order, or sequence of the components are not limited by the terms. Throughout the specification, when a part is said to "include" or "have" a component, this does not mean that other components are excluded, but rather that other components can be further included, unless specifically stated otherwise. In addition, terms such as "part" and "module" described in the specification mean a unit that processes at least one function or operation, and this can be implemented by hardware, software, or a combination of hardware and software.

[0029]

[0030] FIG. 1 illustrates an exemplary environment in which a sleep monitoring device according to some embodiments of the present disclosure may be applied. A system including a user terminal (100), a sleep monitoring device (200), and a sleep sensor (300) illustrated in FIG. 1 can determine a user's sleeping posture using a deep learning model and provide information about the user's sleeping posture to the user terminal (100).

[0031] Below, the operations of the components illustrated in Fig. 1 through the above-described system will be described in more detail.

[0032] FIG. 1 illustrates an example in which a user terminal (100), a sleep monitoring device (200), and a sleep sensor (300) are connected through a network, but this is only for convenience of understanding, and the number of devices that can be connected to the network may vary.

[0033] Meanwhile, Fig. 1 merely illustrates a preferred embodiment for achieving the purpose of the present disclosure, and some components may be added or deleted as needed. Below, the components illustrated in Fig. 1 will be described in more detail.

[0034] The sleep monitoring device (200) can collect and analyze data generated from the user terminal (100). The various information may include all data generated from the user terminal (100). Additionally, this information may be data acquired through a series of electronic devices installed in the user terminal (100).

[0035] Meanwhile, the sleep monitoring device (200) may be implemented with one or more computing devices. For example, all functions of the sleep monitoring device (200) may be implemented with a single computing device. As another example, the first function of the sleep monitoring device (200) may be implemented with a first computing device, and the second function may be implemented with a second computing device. Here, the computing device may be, but is not limited to, a notebook, desktop, laptop, etc., and may include any type of device equipped with computing functions. However, the sleep monitoring device (200) may preferably be implemented with a high-performance, server-level computing device. An example of a computing device will be described with reference to FIG. 5.

[0036] Additionally, the user terminal (100) may be implemented as one or more computing devices, wherein the computing devices may be, but are not limited to, a notebook, desktop, laptop, etc., and may include any type of device equipped with computing capabilities. Additionally, it goes without saying that the user terminal (100) may include a smartphone capable of networking functions.

[0037] In addition, the sleep sensor (300) may include a pressure sensor, a respiration rate sensor, a heart rate sensor, etc., and the sleep sensor (300) may collect various information and data while the user is sleeping and provide the same to the sleep monitoring device (200) or the user terminal (100).

[0038] In addition, the sleep sensor (300) may include a pressure mat, and it is obvious that the sleep sensor (300) may include any electronic device capable of obtaining the user's bio-signal data generated when the user sleeps.

[0039] To avoid redundant description, the various operations performed by the sleep monitoring device (200) will be described in more detail later with reference to the drawings below in FIG. 2.

[0040] In addition, the functions that can be implemented in the sleep monitoring device (200) may also be implemented by utilizing the electronic device mounted on the user terminal (100). Accordingly, in FIG. 1, the sleep monitoring device (200), the user terminal (100), and the sleep sensor (300) are illustrated separately, but according to one embodiment, the sleep monitoring device (200) may be mounted on the user terminal (100) so that the sleep monitoring device (200) may implement the first function, the second function, etc. within the user terminal (100). Therefore, it should be noted that the present invention is not limited to the embodiment in which the user terminal (100) and the sleep monitoring device (200) are externally separated as illustrated in FIG. 1.

[0041] In this specification, for the convenience of explanation, a situation in which a user terminal (100), a sleep monitoring device (200), and a sleep sensor (300) are distinguished and implement functions will be described.

[0042] In some embodiments, the user terminal (100), the sleep monitoring device (200), and the sleep sensor (300), which are components included in the environment to which the sleep monitoring device (200) is applied, may communicate via a network. The network may be implemented as any type of wired / wireless network, such as a local area network (LAN), a wide area network (WAN), a mobile radio communication network, or Wibro (Wireless Broadband Internet).

[0043] Meanwhile, the environment illustrated in FIG. 1 illustrates a connection via a network via a user terminal (100) and a sleep monitoring device (200), but it should be noted that the scope of the present disclosure is not limited thereto, and the user terminal (100), the sleep monitoring device (200), and the sleep sensor (300) may also be connected via P2P (Peer to Peer).

[0044] With reference to FIG. 1, exemplary environments in which a sleep monitoring device (200) according to some embodiments of the present disclosure may be applied have been described. Hereinafter, with reference to the drawings of FIG. 2 and below, methods according to various embodiments of the present disclosure will be described in detail.

[0045] Each step of the methods described below may be performed by a computing device. In other words, each step of the methods may be implemented by one or more instructions executed by a processor of the computing device. All steps included in these methods may be performed by a single physical computing device, but first steps of the method may be performed by a first computing device, and second steps of the method may be performed by a second computing device.

[0046] In the following FIG. 2, the description will continue assuming that each step of the methods is performed by the sleep monitoring device (200) illustrated in FIG. 1. However, for convenience of explanation, the description of the operating entity of each step included in the methods may be omitted.

[0047]

[0048] FIG. 2 is a flowchart illustrating an operation of a sleep monitoring device for determining a user's sleeping posture according to some embodiments of the present disclosure.

[0049] In step S100, the sleep monitoring device (200) may receive sensor data from the sleep sensor (300). The sensor data may be sensor data received at a rate of 1 Hz from a pressure sensor, which is the sleep sensor (300). In step S200, the sleep monitoring device (200) may input the sensor data into a pre-trained deep learning model. At this time, the deep learning model may be trained based on the labeled data by labeling sensor data for the user's sleeping posture as follows: a blank situation, a situation of lying on the left side, a situation of lying on the left side, a situation of lying on the right side, a situation of lying on the right side, a situation of lying upright, a situation of lying with arms crossed, a situation of lying completely on the stomach, a situation of lying on the stomach and using a cell phone, and a situation of lying and using a cell phone. In addition, the deep learning model described in the present invention is a supervised learning algorithm model, and may be a deep learning model that has been previously learned through the labeled learning data, and when sensor data is input into the previously learned deep learning model, the deep learning model can output output data corresponding to various situations described above.

[0050] In addition, the learning data may be labeling data generated from sensor data input through respiration rate and heart rate sensors. For example, the learning data may be sensor data received by the respiration rate and heart rate sensors at a rate of 30 Hz, and may be learning data generated by combining the sensor data with pressure sensor data corresponding to the various postures of the user described above. In this case, the respiration rate may be data generated by storing the average value of the data collected each time and passing through an average filter with a kernel of 60 three times, and may be data that treats the transition point between inhalation and exhalation as respiration rate 1. In addition, in the case of the heart rate, it may be data that passes through an average filter with a kernel of 20 once and is updated every 10 seconds after 60 seconds. When generating the learning data, it may be generated by utilizing sensor data collected for 30 seconds for each posture.

[0051] Accordingly, in step S300, the sleep monitoring device (200) can determine the user's sleeping posture based on the output data output from the pre-trained deep learning model. That is, when sensor data is input, the sleep monitoring device (200) outputs output data from the pre-trained deep learning model, and can determine which posture is labeled in the output data. At this time, the output data is output as a probability value, and for example, if the probability value is calculated as 0.7 for a situation where there is nothing, a threshold probability value is set, and if it is higher than the threshold probability value, the output result is determined as a situation where there is nothing, and information can be provided.

[0052] In step S400, the sleep monitoring device (200) can measure sleep time and non-sleep time according to the sleeping posture. If the user's sleeping posture is determined in step S300, the sleep monitoring device (200) can measure the user's sleep time and non-sleep time according to the sleeping posture and store them in a database.

[0053] In step S500, the sleep monitoring device (200) can provide sleep posture information to the user terminal (100). The sleep monitoring device (200) can generate and provide information on sleep posture, sleep time, and non-sleep time to the user terminal (100). Accordingly, the user can receive information on his / her sleep posture, sleep time, and non-sleep time, and can check the average sleeping posture over a certain period of time and the average sleeping time for each posture. Hereinafter, another embodiment will be described in detail with reference to FIG. 3.

[0054]

[0055] FIG. 3 is a flowchart for a specific description of steps for providing sleep posture information to a user terminal according to some embodiments of the present disclosure.

[0056] In step S510, the sleep monitoring device (200) may receive feedback information regarding sleep posture information from the user terminal (100). The sleep monitoring device (200) may provide information regarding sleep posture to the user terminal (100) and, accordingly, receive feedback information from the user. At this time, the feedback information may be information in which the user inputs inaccurate feedback if the user does not believe the provided sleep posture is accurate, or in which the user inputs accurate feedback if the user believes the provided sleep posture is accurate.

[0057] In step S520, the sleep monitoring device (200) can set deep learning model learning weights for feedback information. Based on the feedback information, if inaccurate feedback information is input, the sleep monitoring device (200) can set the learning weight for the sleep posture to be low in the deep learning model learning weights. In addition, if accurate feedback information is received, the sleep monitoring device (200) can set the learning weight for the sleep posture to be high in the deep learning model learning weights. In step S530, the sleep monitoring device (200) can learn the deep learning model according to the learning weights. Therefore, the sleep monitoring device (200) can utilize the user's feedback information to learn the deep learning model, thereby utilizing the deep learning model more dynamically.

[0058] Below, an example architecture of the deep learning model described in the present invention will be described in detail through FIG. 4.

[0059]

[0060] FIG. 4 is an exemplary diagram of a deep learning model architecture of the present invention according to some embodiments of the present disclosure.

[0061] The deep learning model discussed in the present invention may include all deep learning architectures that can be utilized in supervised learning algorithms, including CNN, a convolutional neural network, RNN, DNN, and LSTM.

[0062] In addition, the artificial neural network, which is a convolutional neural network discussed in the present invention, may include a convolutional layer, a pooling layer, a fully connected layer, and an activating function layer. Here, the convolutional layer includes at least one filter, and at least one filter may have a feature for a specific shape. For example, an edge filter for identifying the outline of an object may have a feature for a specific shape, such as a line that can cover an image. When the edge filter is applied to an image, feature information corresponding to the edge filter can be obtained. A feature map can be extracted through at least one filter included in the convolutional layer.

[0063] Pooling layers can be structurally positioned between convolutional layers. Alternatively, they can be hierarchically positioned after multiple convolutional layers. Pooling layers extract specific values ​​for specific regions from the feature maps extracted through the convolutional layers. Pooling layers come in various types, including max pooling layers and median pooling layers.

[0064] Additionally, a fully connected layer can perform image classification on feature maps extracted through at least one convolutional layer and a pooling layer. For example, a fully connected layer can classify an image by arranging feature maps in a row and then passing them through a hidden layer.

[0065] The Activation Function layer, as an activation function layer, can be applied to feature maps. The Activation Function layer can perform the function of converting quantitative values ​​after the Fully Connected Layer into results indicating whether or not they contain feature information.

[0066] An artificial neural network can learn through training data. The artificial neural network can input training data, output output data, and learn by comparing the training data with the output data (230). For example, if the training data includes label data, the output data may be data including probability values ​​for the label data. The artificial neural network can learn using the error backpropagation method. The error backpropagation method may be a learning method that reflects the errors between the training data and the output data back to the nodes included in the artificial neural network. The artificial neural network can learn in a direction that minimizes the errors between the training data and the output data. When an image is input, the trained artificial neural network can detect objects within the image or classify the type of object. The trained artificial neural network can classify objects based on the properties of the objects within the image.

[0067] This is a general description of the artificial neural network (CNN), i.e., the convolutional neural network, among deep learning models, and is not limited to this interpretation. Referring to Figure 4, one can see an example of the architecture of the deep learning model used in the present invention.

[0068] Referring to Fig. 4, the deep learning model may include a convolutional layer (hereinafter referred to as a convolutional layer) and may be composed of a BNResidual Blocks layer and an SPP layer. At this time, according to Fig. 4, the deep learning model may be structured so that input data may be input to the first convolutional layer, the second convolutional layer, and then pass through the BNResidual Blocks layer. At this time, the first convolutional layer may be 3x3, have a filter size of 8, and have a slide criterion of 1. The second convolutional layer may be 3x3, have a filter size of 16, and have a slide criterion of 1. The BNResidual Blocks layer may have a filter size of 16.

[0069] After the BNResidual Blocks layer, the third convolution layer, the BNResidual Blocks layer, the fourth convolution layer, and the BNResidual Blocks layer may be passed. At this time, the third convolution layer may be 3x3, the filter size may be 32, and the slide criteria may be 1, and the fourth convolution layer may be 3x3, the filter size may be 64, and the slide criteria may be 1. The BNResidual Blocks layer may be 32 and 64 in sequence.

[0070] After this, it can go through the SPP layer, the 5th convolution layer, the 6th convolution layer, and then the GlobalAveragePooling layer. At this time, the 5th convolution layer can be 3x3, 128 in size, and 1 in slide basis, and the 6th convolution layer can be 1x1 10 in size, and 1 in slide basis.

[0071] The BNResidual Blocks layer described in the present invention may be composed of four convolution layers, and at this time, the four convolution layers may sequentially have pixel sizes of 1x1, 3x3, 1x1, and 3x3.

[0072] In addition, the SPP layer described in the present invention includes one convolution layer, and the convolution layer has a size of 1x1 pixel, and then can pass through a Max pooling layer in parallel, and at this time, there can be at least three Max pooling layers, and each can have a size of 5x5, 9x9, or 13x13 pixels. After this, it passes through one convolution layer, which can have a size of 1x1. It should be noted that the architecture of the deep learning model described in FIG. 4 is an example described above, and is not limited thereto, and additionally, a convolution layer, a BNResidual Blocks layer, and an SPP layer can be implemented.

[0073] Below, a system in which a sleep monitoring device (200) can be implemented will be described through FIG. 5.

[0074]

[0075] FIG. 5 is a diagram of an exemplary computing device that may implement a device and / or system according to various embodiments of the present disclosure.

[0076] A computing device (1500) may include one or more processors (1510), a bus (1550), a communication interface (1570), a memory (1530) for loading a computer program (1591) to be executed by the processor (1510), and a storage (1590) for storing the computer program (1591). However, only components related to the embodiment of the present disclosure are illustrated in FIG. 5. Therefore, a person skilled in the art to which the present disclosure pertains may recognize that other general components may be included in addition to the components illustrated in FIG. 5.

[0077] The processor (1510) controls the overall operation of each component of the computing device (1500). The processor (1510) may be configured to include a CPU (Central Processing Unit), an MPU (Micro Processor Unit), an MCU (Micro Controller Unit), a GPU (Graphics Processing Unit), or any other type of processor well known in the art of the present disclosure. In addition, the processor (1510) may perform operations for at least one application or program for executing a method according to embodiments of the present disclosure. The computing device (1500) may include one or more processors.

[0078] The memory (1530) stores various data, commands, and / or information. The memory (1530) may load one or more programs (1591) from the storage (1590) to execute a method according to embodiments of the present disclosure. The memory (1530) may be implemented as a volatile memory such as RAM, but the technical scope of the present disclosure is not limited thereto.

[0079] The bus (1550) provides communication between components of the computing device (1500). The bus (1550) may be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

[0080] The communication interface (1570) supports wired and wireless Internet communication of the computing device (1500). Furthermore, the communication interface (1570) may support various communication methods other than Internet communication. To this end, the communication interface (1570) may be configured to include a communication module well known in the technical field of the present disclosure.

[0081] According to some embodiments, the communication interface (1570) may be omitted.

[0082] Storage (1590) can non-temporarily store one or more programs (1591) and various data.

[0083] Storage (1590) may be configured to include non-volatile memory such as Read Only Memory (ROM), Erasable Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), flash memory, a hard disk, a removable disk, or any form of computer-readable recording medium well known in the art to which the present disclosure pertains.

[0084] The computer program (1591) may include one or more instructions that, when loaded into the memory (1530), cause the processor (1510) to perform methods / operations according to various embodiments of the present disclosure. That is, the processor (1510) may perform the methods / operations according to various embodiments of the present disclosure by executing the one or more instructions.

[0085] Various embodiments of the present disclosure and effects according to the embodiments have been described with reference to FIGS. 1 through 5 so far. The effects according to the technical concept of the present disclosure are not limited to the effects described above, and other effects not mentioned will be clearly understood by those skilled in the art from the description in the specification.

[0086] The technical idea of ​​the present disclosure described with reference to FIGS. 1 to 5 so far can be implemented as a computer-readable code on a computer-readable medium. The computer-readable recording medium can be, for example, a removable recording medium (CD, DVD, Blu-ray disc, USB storage device, removable hard disk) or a fixed recording medium (ROM, RAM, computer-attached hard disk). The computer program recorded on the computer-readable recording medium can be transmitted to another computing device via a network such as the Internet and installed on the other computing device, thereby allowing it to be used on the other computing device.

[0087] Although all components constituting the embodiments of the present disclosure have been described as being combined or operated in combination as one, the technical concept of the present disclosure is not necessarily limited to such embodiments. That is, within the scope of the present disclosure, all components may be selectively combined and operated one or more times.

[0088] Although operations are depicted in the drawings in a particular order, this should not be understood to imply that the operations must be performed in the particular order depicted, or in any sequential order, or that all depicted operations must be performed to achieve the desired results. In certain circumstances, multitasking and parallel processing may be advantageous. Furthermore, the separation of the various components in the embodiments described above should not be understood to imply that such separation is absolutely necessary, and it should be understood that the program components and systems described may generally be integrated together into a single software product or packaged into multiple software products.

[0089] Although the embodiments of the present disclosure have been described with reference to the attached drawings, those skilled in the art will appreciate that the present disclosure can be implemented in other specific forms without changing the technical idea or essential features thereof. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. The scope of protection of the present disclosure should be interpreted by the claims below, and all technical ideas within a scope equivalent thereto should be interpreted as being included in the scope of rights of the technical ideas defined by the present disclosure.

Claims

1. A method performed by a sleep monitoring device, A step of receiving sensor data from a sleep sensor; A step of inputting sensor data into a trained deep learning model; A step of determining the user's sleeping posture based on output data from a trained deep learning model; Step of measuring sleep time and non-sleep time according to sleep position; and A step of providing sleep posture information to a user terminal; comprising: A method for determining a user's sleeping position and providing sleeping position information using a deep learning model.

2. In paragraph 1, The step of providing sleep posture information to the user terminal further includes the step of receiving feedback information about sleep posture from the user terminal. A method for determining a user's sleeping position and providing sleeping position information using a deep learning model.

3. In paragraph 2, The step of providing sleep posture information to the user terminal further includes the step of setting deep learning model learning weights for feedback information. A method for determining a user's sleeping position and providing sleeping position information using a deep learning model.

4. In paragraph 3, The step of providing sleep posture information to the user terminal further includes a step of learning a deep learning model according to learning weights. A method for determining a user's sleeping position and providing sleeping position information using a deep learning model.

5. In paragraph 1, The above deep learning model is characterized by including at least one convolutional layer, a BNResidual Blocks layer, and an SPP layer. A method for determining a user's sleeping position and providing sleeping position information using a deep learning model.

6. In paragraph 5, The BNResidual Blocks layer in the above deep learning model consists of at least one convolutional layer. A method for determining a user's sleeping position and providing sleeping position information using a deep learning model.

7. In paragraph 6, The SPP layer in the above deep learning model is composed of at least one convolutional layer and at least one max pooling layer. A method for determining a user's sleeping position and providing sleeping position information using a deep learning model.

8. Processor; network interface; memory; and A computer program loaded into the above memory and executed by the above processor, The above processor, Instructions for receiving sensor data from a sleep sensor; Instructions for inputting sensor data into a trained deep learning model; An instruction for determining the user's sleeping posture based on output data from a trained deep learning model; Instructions for measuring sleep time and non-sleep time according to sleep position; and Instructions for providing sleep posture information to a user terminal; including performing A sleep monitoring device that uses a deep learning model to determine the user's sleeping position and provide sleeping position information.

9. In paragraph 8, The instruction for providing sleep posture information to a user terminal further includes an instruction for receiving feedback information about sleep posture from the user terminal. A sleep monitoring device that uses a deep learning model to determine the user's sleeping position and provide sleeping position information.

10. In paragraph 9, The instruction for providing sleep posture information to the user terminal further includes an instruction for setting deep learning model learning weights for feedback information. A sleep monitoring device that uses a deep learning model to determine the user's sleeping position and provide sleeping position information.

11. In paragraph 10, The instruction for providing sleep posture information to the user terminal further includes an instruction for learning a deep learning model according to learning weights. A sleep monitoring device that uses a deep learning model to determine the user's sleeping position and provide sleeping position information.

12. In paragraph 8, The above deep learning model is characterized by including at least one convolutional layer, a BNResidual Blocks layer, and an SPP layer. A sleep monitoring device that uses a deep learning model to determine the user's sleeping position and provide sleeping position information.

13. In paragraph 12, The BNResidual Blocks layer in the above deep learning model consists of at least one convolutional layer. A sleep monitoring device that uses a deep learning model to determine the user's sleeping position and provide sleeping position information.

14. In paragraph 13, The SPP layer in the above deep learning model is composed of at least one convolutional layer and at least one max pooling layer. A sleep monitoring device that uses a deep learning model to determine the user's sleeping position and provide sleeping position information.

Images