Wearable device control method and device, equipment, medium and program product

Abstract

The invention relates to a wearable equipment control method and device, equipment, a medium and a program product, the wearable equipment control method comprises the steps that taking-off and wearing action information is determined, the taking-off and wearing action information is used for representing taking-off and wearing action types, and the taking-off and wearing action types comprise a wearing action, a taking-off action and a non-taking-off and non-wearing action; and performing living body identification detection in response to the wearing action or the removing action of the wearing action type. According to the embodiment of the invention, the wearing action and the removing action are taken as the triggering conditions of the living body recognition detection, so that the living body recognition detection is only carried out when the wearing and removing action of the wearable device occurs, the triggering of the living body recognition detection can be adapted to the detection requirement, the detection triggering rate and the interference to the detection result are reduced, and the user experience is improved.

Description

Technical Field

[0001] This disclosure relates to the field of wear detection, specifically to a wearable device control method, apparatus, equipment, medium, and program product. Background Technology

[0002] In recent years, with the rapid development of physiological indicator detection technology and the continuous iteration of wearable devices, wearable devices with physiological indicator detection functions such as blood oxygen and heart rate have been widely used in people's daily health monitoring. To reduce the power consumption of wearable devices and improve the accuracy of physiological indicator detection, it is necessary to determine the wearing status of wearable devices through liveness detection.

[0003] However, the use of related technologies for liveness detection has resulted in loose triggering conditions, leading to problems such as excessively high trigger rates and easily interfered detection results, resulting in a poor user experience. Summary of the Invention

[0004] To overcome the problems existing in related technologies, this disclosure provides a wearable device control method, apparatus, device, medium, and program product.

[0005] According to a first aspect of the present disclosure, a wearable device control method is provided, the wearable device control method comprising:

[0006] Determine the information on wearing and removing actions, which is used to characterize the type of wearing and removing actions, including wearing actions, removing actions, and no wearing / removing actions;

[0007] In response to the action type being either wearing or removing, a liveness detection is performed.

[0008] In some embodiments of this disclosure, determining the wearing / removing action information includes:

[0009] Acquire the fixed infrared signal within the first preset time period before the current time;

[0010] The information regarding the putting on / taking off action is determined based on the fixed infrared signal.

[0011] In some embodiments of this disclosure, acquiring the fixed infrared signal within a first preset time period prior to the current moment includes:

[0012] Continuously acquire acceleration signals and the fixed infrared signals;

[0013] Based on the acceleration signal, action level information is determined, which is used to characterize the action level within a second preset time period before the current moment.

[0014] In response to the action level being greater than a preset action level, a fixed infrared signal within the first preset time period prior to the current moment is acquired.

[0015] In some embodiments of this disclosure, determining the action level information based on the acceleration signal includes:

[0016] The acceleration signal within the second preset time period before the current moment is filtered to obtain a filtered signal;

[0017] The action level information is determined based on the filtered signal.

[0018] In some embodiments of this disclosure, the step of acquiring the fixed infrared signal within a first preset time period prior to the current moment further includes:

[0019] Based on the acceleration signal, action cycle information is determined, which is used to characterize whether the action within a third preset time period after the current moment is periodic.

[0020] The step of responding to an action level greater than a preset action level by acquiring a fixed infrared signal within the first preset time period prior to the current moment includes:

[0021] In response to the action level being greater than a preset action level and the action not being periodic, a fixed infrared signal within the first preset time period prior to the current moment is acquired.

[0022] In some embodiments of this disclosure, determining the motion cycle information based on the acceleration signal includes:

[0023] Calculate the main frequency information of the acceleration signal within multiple fourth preset time periods. The termination times of the multiple fourth preset time periods are set at intervals within a third preset time period after the current time. The fourth preset time periods are longer than the third preset time periods.

[0024] Based on the aforementioned main frequency information, the action cycle information is determined.

[0025] In some embodiments of this disclosure, determining the donning / removing action information based on the fixed infrared signal includes:

[0026] The fixed infrared signal is preprocessed to obtain a preprocessed signal;

[0027] The preprocessed signal is input into the donning and wearing action information determination model, and the donning and wearing action information is determined based on the output of the donning and wearing action information determination model.

[0028] In some embodiments of this disclosure, there are multiple fixed infrared signals, and the model for determining the wearing / removing action information includes a first model and a second model. The first model is capable of linear feature recognition, and the second model is capable of nonlinear feature recognition.

[0029] The preprocessing of the fixed infrared signal to obtain a preprocessed signal includes:

[0030] Each of the fixed infrared signals is preprocessed to obtain a preprocessed signal that corresponds one-to-one with each of the fixed infrared signals.

[0031] The step of inputting the preprocessed signal into the donning / removal action information determination model, and determining the donning / removal action information based on the output of the donning / removal action information determination model, includes:

[0032] The preprocessed signals are respectively input into the first model and the second model to obtain the first output result of the first model and the second output result of the second model;

[0033] The first output result and the second output result are fused together to determine the wearing and removing action information.

[0034] In some embodiments of this disclosure, the wearable device control method further includes:

[0035] Based on the ambient light signal, ambient light change information is determined, which is used to characterize the degree of change of ambient light within a first preset time period before the current moment.

[0036] The step of determining the wearing / removing action information based on the output of the model includes: determining the wearing / removing action information based on the output and the ambient light change information.

[0037] In some embodiments of this disclosure, the first model includes a Fisher spatiotemporal projection model; and / or,

[0038] The second model includes a spatiotemporal convolution model.

[0039] In some embodiments of this disclosure, the sample data used to train the model for determining the wearing and removing action information is obtained in the following manner:

[0040] The sample data is obtained by extracting infrared signal segments corresponding to each type of wearing and removing action from infrared sample signals that contain multiple wearing and removing actions.

[0041] Each of the wearing actions corresponds to multiple infrared signal segments, and for the same wearing action, the position of the action mark time in different infrared signal segments is different.

[0042] Each removal action corresponds to a mark time in multiple infrared signal segments. For the same removal action, the mark time is located in different infrared signal segments.

[0043] In some embodiments of this disclosure, after performing liveness detection in response to the wearing / removing action being either a wearing action or a removal action, the wearable device control method further includes:

[0044] When the type of the wearing and removing action is a wearing action, in response to the detection result of the liveness detection being a liveness, the wearing state of the wearable device is determined to be a wrist state;

[0045] When the type of wearing action is removal, in response to the detection result of the liveness detection being non-live, the wearing state of the wearable device is determined to be wrist-free.

[0046] According to a second aspect of the present disclosure, a wearable device control device is provided, the wearable device control device comprising:

[0047] A determining module is used to determine the wearing / removing action information, which is used to characterize the wearing / removing action type, including wearing action, removal action, and no wearing / removing action.

[0048] The detection module is used to perform liveness detection in response to whether the wearing or removing action is a wearing action or a removing action.

[0049] According to a third aspect of the present disclosure, an electronic device is provided, the electronic device comprising:

[0050] processor;

[0051] Memory used to store processor-executable instructions;

[0052] The processor is configured to execute the wearable device control method as described in the first aspect.

[0053] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium is provided, wherein when instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the wearable device control method as described in the first aspect.

[0054] According to a fifth aspect of the present disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the wearable device control method as described in the first aspect.

[0055] The technical solutions provided by the embodiments of this disclosure may include the following beneficial effects: using wearing and removing actions as triggering conditions for liveness detection, so that liveness detection is only performed when the wearable device is worn or removed, ensuring that the triggering of liveness detection can be adapted to the detection requirements, reducing the detection trigger rate and interference with the detection results, and improving the user experience.

[0056] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0057] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0058] Figure 1 This is a flowchart illustrating a wearable device control method according to an exemplary embodiment.

[0059] Figure 2 This is a flowchart illustrating the determination of information for putting on and taking off clothing, according to an exemplary embodiment.

[0060] Figure 3 This is a schematic diagram illustrating a fixed infrared signal according to an exemplary embodiment.

[0061] Figure 4 This is a schematic diagram illustrating a fixed infrared signal according to another exemplary embodiment.

[0062] Figure 5 This is a schematic diagram illustrating a fixed infrared signal according to another exemplary embodiment.

[0063] Figure 6 This is a flowchart illustrating, according to an exemplary embodiment, the acquisition of a fixed infrared signal within a first preset time period prior to the current moment.

[0064] Figure 7 This is a flowchart illustrating, according to an exemplary embodiment, the determination of motion level information based on an acceleration signal.

[0065] Figure 8 This is a flowchart illustrating, according to an exemplary embodiment, the determination of motion cycle information based on an acceleration signal.

[0066] Figure 9 This is a flowchart illustrating, according to an exemplary embodiment, the determination of wearing / removing action information based on a fixed infrared signal.

[0067] Figure 10This is a flowchart illustrating, according to an exemplary embodiment, a process of inputting a preprocessed signal into a model for determining wearing and removing motion information, and determining wearing and removing motion information based on the output of the model.

[0068] Figure 11 This is a schematic diagram illustrating an ambient light signal according to an exemplary embodiment.

[0069] Figure 12 This is a schematic diagram of the Fisher spacetime projection model according to an exemplary embodiment.

[0070] Figure 13 This is a schematic diagram illustrating a spatiotemporal convolution model according to an exemplary embodiment.

[0071] Figure 14 This is a flowchart illustrating a wearable device control method according to another exemplary embodiment.

[0072] Figure 15 This is a flowchart illustrating a wearable device control method according to another exemplary embodiment.

[0073] Figure 16 This is a block diagram illustrating a wearable device control device according to an exemplary embodiment.

[0074] Figure 17 This is a block diagram of an electronic device according to an exemplary embodiment.

[0075] In the picture:

[0076] 10-Determining module; 20-Detection module; 101-Processing component; 102-Memory; 103-Power component; 104-Multimedia component; 105-Audio component; 106-Input / output interface; 107-Sensor component; 108-Communication component; 109-Processor. Detailed Implementation

[0077] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention as detailed in the appended claims.

[0078] In recent years, with the rapid development of physiological indicator detection technology and the continuous iteration of wearable devices, such as watches and wristbands that display physiological indicators like blood oxygen and heart rate, wearable devices have been widely used for health monitoring in people's daily lives. To reduce power consumption and improve the accuracy of physiological indicators, it is necessary to determine the wearing status of wearable devices through liveness detection. This allows for the cessation of real-time monitoring of various physiological indicators when the liveness detection result indicates a non-live person, i.e., the wearing status is wrist-free.

[0079] In related technologies, the presence of an obstruction is determined by the magnitude of the infrared DC value, triggering liveness detection when an obstruction is present. However, the triggering condition of obstruction in these technologies is extremely lenient. For example, non-living objects such as clothing, pendants, and bags can also obstruct the wearable device, leading to numerous unnecessary scenarios that can trigger liveness detection and wasting power. Furthermore, after a non-living object triggers liveness detection, the rhythmic signal returned from the cavity between the non-living object and the wearable device is similar to that of a living person. This can easily lead to the non-living object being identified as a living person, interfering with the detection results and affecting the accuracy of the wearable device's wearing status, resulting in a poor user experience.

[0080] Based on this, an exemplary embodiment of this disclosure provides a wearable device control method. By determining the wearing / removing action information and performing liveness detection when the wearing / removing action type represented by the information is either a wearing action or a removal action, adaptive triggering of liveness detection is achieved, providing a basis for determining the wearing state. By using wearing and removing actions as triggering conditions for liveness detection, liveness detection is performed only when the wearable device is worn / removed, ensuring that the triggering of liveness detection is compatible with detection requirements, reducing the detection trigger rate and interference with detection results, and improving the user experience.

[0081] In one exemplary embodiment, a wearable device control method is provided, applied to a wearable device, which may include, for example, a watch, a bracelet, or other device that can be worn on a specific part of a user's body and has physiological indicator detection functions. (Reference) Figure 1 As shown, the wearable device control method includes:

[0082] S100. Determine the information on wearing and removing actions. The information on wearing and removing actions is used to characterize the type of wearing and removing actions. The types of wearing and removing actions include wearing actions, removing actions, and no wearing or removing actions.

[0083] In step S100, the information on wearing and removing the device is determined. This information characterizes the type of wearing and removing action, which includes wearing, removing, and no wearing / removing action. The wearable device can belong to any of these three types. Wearing indicates that the user has worn the wearable device. After wearing, it can be preliminarily determined that the wearable device has switched from a wrist-free state to a wrist-on state, at which point liveness detection is needed to further determine the wearing status. Removing indicates that the user has removed the wearable device. After removing, it can be preliminarily determined that the wearable device has switched from a wrist-on state to a wrist-free state, at which point liveness detection is needed to further determine the wearing status. No wearing / removing action indicates that the user has not performed any wearing or removing action, and it can be preliminarily determined that the wearing status of the wearable device has not changed; therefore, liveness detection is not needed.

[0084] For example, a fixed infrared signal with a fixed emission current can be used as the basis for determining the wearing and removing action information. The type of wearing and removing action can be determined by identifying the change pattern of the fixed infrared signal or extracting its features. Alternatively, acceleration signals, gyroscope signals, temperature signals, impedance signals, etc., can be used as the basis for determining the wearing and removing action information. The type of wearing and removing action can be determined by identifying the change pattern of the signal or extracting its features.

[0085] S200: In response to the action type of wearing or removing, perform liveness detection.

[0086] In step S200, as mentioned above, when the type of wearing action is wearing or removing, there is a need to perform liveness detection to further determine the wearing status. Therefore, when the type of wearing action represented by the wearing action information is wearing or removing, liveness detection is performed. When the type of wearing action represented by the wearing action information is no wearing action, liveness detection is kept off.

[0087] For example, liveness detection can rely on multimodal signals to extract liveness features, using information such as heart rate, pulse wave, and variance to determine whether a live person is within the detection range of the photoplethysmography (PPG) sensor. After liveness detection, the wearing status of the wearable device can be determined based on the detection results, and the need to detect physiological indicators such as blood oxygen and heart rate can be determined based on the wearing status of the wearable device.

[0088] In this embodiment, by determining the wearing and removing action information and performing liveness detection when the wearing and removing action type represented by the information is either wearing or removing, adaptive triggering of liveness detection is achieved, providing a basis for determining the wearing status. Using wearing and removing actions as trigger conditions for liveness detection ensures that liveness detection only occurs when the wearable device is worn or removed, guaranteeing that the triggering of liveness detection is compatible with detection requirements, reducing the detection trigger rate and interference with detection results, and improving the user experience.

[0089] In some embodiments, reference Figure 2 As shown, the information for determining the wearing and removing actions includes:

[0090] S110. Obtain the fixed infrared signal within the first preset time period before the current time.

[0091] In step S110, a fixed infrared signal within a first preset time period prior to the current time is acquired. The first preset time period can be set according to detection requirements and test results, and can be any duration within the range of 3s to 5s. The fixed infrared signal can be acquired by an infrared sensor. The fixed infrared signal is an infrared signal with a fixed emission current, so that the fixed infrared signal has a fixed infrared light intensity. The number of fixed infrared signals can be one or more, that is, a single-channel fixed infrared signal within the first preset time period prior to the current time can be acquired, or multiple-channel fixed infrared signals within the first preset time period prior to the current time can be acquired simultaneously.

[0092] For example, a first preset time period can be used as the window length of the sliding window, and the sliding window can be slid with a specific step size in the continuous fixed infrared signal. Acquisition conditions can be set for the acquisition of the fixed infrared signal. When the acquisition conditions are met, the fixed infrared signal in the corresponding sliding window can be used as the fixed infrared signal in the first preset time period before the current time.

[0093] S120. Determine the wearing / removing action information based on the fixed infrared signal.

[0094] In step S120, the fixed infrared signal has specific characteristics and change patterns under different wearing and removing actions, that is, the fixed infrared signal can have different signal performance under different wearing and removing actions. Figures 3 to 5 The figure shows four fixed infrared signals for different testers. The fixed infrared signals under the wearing and removing actions have specific signal characteristics, and each wearing or removing action is consistent in temporal morphology. Therefore, the fixed infrared signals can be used as a basis for judging the type of wearing and removing action. The wearing and removing action information can be determined based on the fixed infrared signals within a first preset time period before the current time.

[0095] In this embodiment, by acquiring a fixed infrared signal within a first preset time period prior to the current moment, and determining the wearing / removal action information based on the fixed infrared signal, the wearing / removal action information is determined, providing a basis for whether to activate liveness detection. Using the fixed infrared signal as the basis for determining the wearing / removal action information fully utilizes the specific signal characteristics of the fixed infrared signal under different wearing / removal actions, ensuring the accuracy of judging the type of wearing / removal action, thereby improving the accuracy of triggering liveness detection.

[0096] In some embodiments, reference Figure 6 As shown, the method for acquiring a fixed infrared signal within a first preset time period prior to the current moment includes:

[0097] S111, continuously acquire acceleration signals and fixed infrared signals.

[0098] In step S111, the wearable device continuously collects acceleration signals and fixed infrared signals. The acceleration signals may include, for example, triaxial acceleration signals. The acceleration signals are acquired by the acceleration sensor of the wearable device, and the sampling frequency of the acceleration signals may be, for example, 25Hz.

[0099] S112. Based on the acceleration signal, determine the action level information. The action level information is used to characterize the action level within the second preset time period before the current moment.

[0100] In step S112, the acceleration signal can characterize the acceleration change of the wearable device, and the acceleration change represents the generation of action. Therefore, action level information can be determined based on the acceleration signal. The action level information can characterize the action level within a second preset time period before the current moment. The second preset time period can be, for example, 0.4s. The action level can be divided into three levels: small action, medium action, and large action.

[0101] S113. In response to an action level greater than a preset action level, obtain the fixed infrared signal within the first preset time period before the current moment.

[0102] In step S113, the preset action level can be, for example, a medium action. When the action level is greater than the preset action level, it means that a large action has occurred in the second preset time period before the current time. This action is more likely to be a wearing action or a removing action. At this time, it is necessary to determine the wearing and removing action information through fixed infrared signals. Then, the continuously acquired fixed infrared signals are intercepted to obtain the fixed infrared signals in the first preset time period before the current time.

[0103] For example, a second preset time period of 0.4s is used as the window length of the sliding window, and the sliding window is slid in a step of 0.04s in the continuous acceleration signal. The corresponding action level information is determined according to the acceleration signal in each sliding window. When the action level information represents an action greater than the preset action level, the end time of the sliding window is taken as the current time, and the fixed infrared signal 5s before the current time is intercepted to obtain the fixed infrared signal in the first preset time period before the current time, which is used to determine the removal and wearing action information.

[0104] Understandably, acquiring fixed infrared signals within a first preset time period before the current moment when the action level is greater than the preset action level ensures that the fixed infrared signals within the first preset time period only participate in determining the wearing / unwearing action information when a large action occurs. When there is no action or the action is small, the determination of wearing / unwearing action information is not triggered, so that the determination of wearing / unwearing action information can be adapted to actual needs and avoids the power consumption waste caused by continuously determining wearing / unwearing action information.

[0105] In this embodiment, by continuously acquiring acceleration signals and fixed infrared signals, and determining the action level information based on the acceleration signals, when the action level is greater than a preset action level, the fixed infrared signals within a first preset time period before the current moment are acquired. This achieves the acquisition of fixed infrared signals within the first preset time period before the current moment, providing a basis for determining the wearing / removing action information. Using an action level greater than the preset action level as the condition for acquiring fixed infrared signals within the first preset time period before the current moment enables the matching and windowing processing of fixed infrared signals, setting a trigger condition for determining the wearing / removing action information. This ensures that the determination of the wearing / removing action information can be adapted to actual needs, avoiding the power waste caused by continuously determining the wearing / removing action information.

[0106] In some embodiments, reference Figure 7 As shown, based on the acceleration signal, the action level information is determined, including:

[0107] S112-1. Filter the acceleration signal within the second preset time period before the current time to obtain the filtered signal.

[0108] S112-2. Determine the action level information based on the filtered signal.

[0109] In steps S112-1 and S112-2, the acceleration signal within the second preset time period before the current time is filtered to obtain a filtered signal. Then, the magnitude of the action level is quantized based on the filtered signal to determine the action level information.

[0110] For example, the acceleration signal is a triaxial acceleration signal. The acceleration signal within a second preset time period of 0.4s before the current moment can be filtered to smooth the signal and remove noise, resulting in a filtered signal. The filtered signal is then subjected to acceleration modulus calculation in the three axes to characterize the overall acceleration magnitude. Based on the overall acceleration magnitude, the action level within the second preset time period before the current moment is determined, thereby realizing the determination of action level information.

[0111] In this embodiment, the acceleration signal within a second preset time period prior to the current moment is filtered to obtain a filtered signal. The action level information is then determined based on the filtered signal, providing a basis for acquiring the fixed infrared signal within a first preset time period prior to the current moment. By filtering the acceleration signal, signal smoothing and noise removal are achieved, improving the accuracy of determining the action level information.

[0112] In some embodiments, acquiring a fixed infrared signal within a first preset time period prior to the current moment further includes: determining motion cycle information based on an acceleration signal, wherein the motion cycle information is used to characterize whether the motion within a third preset time period after the current moment is periodic. Acquiring a fixed infrared signal within the first preset time period prior to the current moment in response to a motion level greater than a preset motion level includes: acquiring a fixed infrared signal within the first preset time period prior to the current moment in response to a motion level greater than a preset motion level and the motion not being periodic.

[0113] As mentioned earlier, when acquiring the fixed infrared signal within the first preset time period before the current moment, it is necessary to determine the action level information based on the acceleration signal. At this time, the action cycle information can also be determined based on the acceleration signal. The action cycle information can characterize whether the action within the third preset time period after the current moment is periodic. The third preset time period can be, for example, 5 seconds. If the action cycle information indicates that the action within the third preset time period after the current moment is periodic, it means that the action occurring within the second preset time period before the current moment has a high probability of being a periodic action with a specific frequency, such as walking or arm swinging. The possibility of this action being a putting-on or taking-off action is low, and it is not necessary to determine the putting-on / taking-off action information based on the fixed infrared signal within the first preset time period before the current moment. Therefore, the fixed infrared signal within the first preset time period before the current moment is acquired only when the action level is greater than the preset action level and the action level information indicates that the action within the third preset time period after the current moment is not periodic, in order to determine the putting-on / taking-off action information.

[0114] It is understandable that when the action level is greater than the preset action level and the action is not periodic, the fixed infrared signal within the first preset time period before the current moment is obtained. This ensures that the fixed infrared signal within the first preset time period is used to determine the wearing and removing action information only when a large non-periodic action occurs. This ensures that the determination of the wearing and removing action information can be adapted to the actual needs and avoids the power consumption waste caused by determining the wearing and removing action information when a periodic action occurs.

[0115] In this embodiment, the action cycle information is determined based on the acceleration signal. When the action level is greater than a preset action level and the action is not periodic, a fixed infrared signal within a first preset time period before the current moment is acquired, providing a basis for determining the wearing / removing action information. Using both action level information and action cycle information as the conditions for acquiring the fixed infrared signal within the first preset time period before the current moment further limits the triggering conditions for determining the wearing / removing action information, avoiding the power consumption waste caused by determining the wearing / removing action information when periodic actions occur.

[0116] In some embodiments, reference Figure 8 As shown, based on the acceleration signal, the motion cycle information is determined, including:

[0117] S114. Calculate the main frequency information of the acceleration signal within multiple fourth preset time periods. The termination times of the multiple fourth preset time periods are set at intervals within the third preset time period after the current time. The fourth preset time period is longer than the third preset time period.

[0118] In step S114, acceleration signals within multiple fourth preset time periods are extracted, and the dominant frequency information of the acceleration signals within these time periods is calculated. The dominant frequency information is used to characterize the dominant frequency component in the signal, i.e., the most significant signal component, and can be extracted using Fourier transform.

[0119] The fourth preset time period can be, for example, 10 seconds. The end times of multiple fourth preset time periods are set at intervals within the third preset time period, meaning that the interval between two adjacent fourth preset time periods is the same, and the end time of the first fourth preset time period is the current time, while the end time of the last fourth preset time period is the end time of the third preset time period after the current time. The fourth preset time period is longer than the third preset time period, ensuring that the start time of each fourth preset time period is before the current time.

[0120] For example, the window length of the sliding window is set to the fourth preset time period of 10 seconds, and the sliding window is slid in the acceleration signal with a step size of 1 second, so that the end time of the sliding window is slid from the current time to the end time of the third preset time period of 5 seconds after the current time, and the main frequency information of the acceleration signal included at each position of the sliding window is calculated respectively.

[0121] S115. Determine the action cycle information based on the main frequency information.

[0122] In step S115, the main frequency information corresponding to each fourth preset time period is integrated, and the frequency structure of the acceleration signal in each fourth preset time period is used to determine whether the action in the third preset time period after the current moment is periodic, thereby realizing the determination of the action period information.

[0123] In this embodiment, the dominant frequency information of acceleration signals within multiple fourth preset time periods is calculated, and the action cycle information is determined based on each dominant frequency information, providing a basis for acquiring fixed infrared signals within the first preset time period before the current moment. Using the dominant frequency information corresponding to multiple fourth preset time periods as the basis for determining the action cycle information allows for the determination of whether the action is periodic based on the frequency structure of the acceleration signals before and after the current moment, thus improving the accuracy of determining the action cycle information.

[0124] In some embodiments, reference Figure 9 As shown, based on a fixed infrared signal, the information for putting on and taking off the garment is determined, including:

[0125] S121. Preprocess the fixed infrared signal to obtain a preprocessed signal.

[0126] In step S121, the fixed infrared signal acquired within a first preset time period prior to the current moment is preprocessed to obtain a preprocessed signal. Preprocessing of the fixed infrared signal may include, for example, outlier detection and baseline removal. Outlier detection is used to identify and correct errors and noise in the signal to ensure data reliability. Baseline removal is used to remove baseline signals or trends to reveal more meaningful changes or features, playing a crucial role in ensuring signal quality and the accuracy of analysis results.

[0127] S122. Input the preprocessed signal into the removal and wearing action information determination model, and determine the removal and wearing action information based on the output of the removal and wearing action information determination model.

[0128] In step S122, the preprocessed signal is input to the removal and wearing action information determination model. The removal and wearing action information determination model may include, for example, a trained neural network model or a machine learning model. The removal and wearing action information determination model can analyze the input preprocessed signal and output the corresponding results. The removal and wearing action information can be determined based on the output results of the removal and wearing action information determination model.

[0129] In this embodiment, the fixed infrared signal is preprocessed to obtain a preprocessed signal, which is then input into the donning / wearing action information determination model. The donning / wearing action information is determined based on the model's output, thus providing a basis for deciding whether to activate liveness detection. Preprocessing the fixed infrared signal ensures signal quality and accuracy of the analysis results. Using the output of the donning / wearing action information determination model as the basis for determining the donning / wearing action information improves the efficiency and stability of this determination process.

[0130] In some embodiments, there are multiple fixed infrared signals, and the model for determining the wearing and removing action information includes a first model and a second model. The first model is capable of linear feature recognition, and the second model is capable of nonlinear feature recognition.

[0131] There are multiple fixed infrared signals, each of which is a fixed infrared signal within a first preset time period prior to the current time. This ensures that each acquisition captures multiple channels of fixed infrared signals within the same first preset time period. Each fixed infrared signal is acquired by an infrared sensor positioned at a different location. The number of fixed infrared signals, i.e., the number of channels, can be, for example, four.

[0132] The model for determining the wearing / removing action information includes a first model and a second model. The first model can effectively identify linear features from multi-dimensional data, while the second model can effectively identify non-linear features from multi-dimensional data. This allows the first and second models to extract linear and non-linear features from a fixed infrared signal within a first preset time period before the current moment, respectively, and output corresponding results based on the extracted feature types. The first model may include, for example, a Fisher spatiotemporal projection model, a Naive Bayes classifier model, or a support vector machine model, while the second model may include, for example, a spatiotemporal convolution model, a mean clustering model, or a principal component analysis model.

[0133] Preprocessing the fixed infrared signals to obtain preprocessed signals includes: preprocessing each fixed infrared signal separately to obtain a preprocessed signal that corresponds one-to-one with each fixed infrared signal.

[0134] When preprocessing fixed infrared signals, each fixed infrared signal, i.e., the fixed infrared signal of each channel, needs to be preprocessed separately, such as outlier detection and baseline removal, to obtain a preprocessed signal that corresponds one-to-one with each fixed infrared signal, so as to ensure signal quality and accuracy of analysis results.

[0135] refer to Figure 10 As shown, the preprocessed signal is input into the donning / removal action information determination model, and the donning / removal action information is determined based on the output of the model, including:

[0136] S122-1. Input multiple preprocessed signals into the first model and the second model respectively to obtain the first output result of the first model and the second output result of the second model.

[0137] In step S122-1, multiple preprocessed signals corresponding one-to-one with multiple fixed infrared signals are input to the first model and the second model respectively. The linear and nonlinear characteristics of the multiple fixed infrared signals are identified by the first model and the second model respectively, and the first output result output by the first model and the second output result output by the second model are obtained. The first output result and the second output result can respectively characterize the type of putting on and taking off action determined under different models.

[0138] S122-2. Perform a fusion decision on the first and second output results to determine the action information for putting on and taking off the garment.

[0139] In step S122-2, the first output result and the second output result are fused and decided to comprehensively consider the linear and nonlinear characteristics of multiple fixed infrared signals, and the final type of donning and putting on action is determined according to the decision result, thereby realizing the determination of donning and putting on action information and improving the accuracy of donning and putting on action information.

[0140] For example, the first and second output results can be numerical values ​​representing the types of donning and putting on clothing actions, with different numerical ranges corresponding to different donning and putting on clothing action types. The first and second output results can be fused and decided by weighted averaging to obtain a weighted average decision value, and the corresponding donning and putting on clothing action type can be determined according to the numerical range of the decision value, which serves as the final donning and putting on clothing action information.

[0141] In this embodiment, by acquiring multiple fixed infrared signals and using a first model and a second model as models for determining the wearing / removing action information, the linear and nonlinear characteristics of the multiple fixed infrared signals can be identified using the first and second models respectively. Furthermore, the first and second output results corresponding to the first and second models are fused for decision-making, thereby determining the wearing / removing action information. Using both linear and nonlinear characteristics of multiple fixed infrared signals as the basis for determining the wearing / removing action information improves the accuracy and robustness of the determined information.

[0142] In some embodiments, the wearable device control method further includes: determining ambient light change information based on ambient light signals, wherein the ambient light change information is used to characterize the degree of change in ambient light within a first preset time period prior to the current moment; and determining the wearing / removing action information based on the output of the model, including: determining the wearing / removing action information based on the output and the ambient light change information.

[0143] Wearable devices can also be equipped with ambient light sensors, which can acquire ambient light signals and determine ambient light change information based on these signals. This ambient light change information characterizes the degree of change in ambient light within a first preset time period prior to the current moment, that is, the degree of change in ambient light within the same time range as the acquired fixed infrared signal.

[0144] like Figure 11 As shown, when the action of putting on or taking off occurs, the ambient light signal will undergo a drastic change. That is, the ambient light signal can also have a specific signal performance when the action of putting on or taking off occurs. Therefore, the degree of change in ambient light can also be used as a basis for judging the type of action of putting on or taking off. When determining the action information of putting on or taking off, the output result of the model and the ambient light change information can be determined simultaneously based on the action information.

[0145] For example, the ambient light change information can be a value within the range of 0 to 1, with smaller values ​​indicating a lower degree of ambient light change. When the model for determining the wearing / removing action information includes a first model and a second model, after fusing the first and second output results using a weighted average to obtain a weighted average decision value, the ambient light change information can be multiplied by the decision value to adjust the fused decision result. The type of wearing / removing action is then determined based on the numerical range of the adjusted decision value, serving as the final wearing / removing action information.

[0146] In this embodiment, ambient light change information is determined based on the ambient light signal, and the removal / wearing action information is determined based on the output result and the ambient light change information. The ambient light change information and the output result of the removal / wearing action information determination model can be used together as the basis for determining the removal / wearing action information. The judgment of the removal / wearing action type can be adjusted according to the degree of change of ambient light, making full use of the specific signal performance of the ambient light signal under the removal / wearing action, and further improving the accuracy of judging the removal / wearing action type.

[0147] In some embodiments, the first model includes a Fisher spatiotemporal projection model, or the second model includes a spatiotemporal convolution model. Alternatively, the first model may include a Fisher spatiotemporal projection model, and the second model may include a spatiotemporal convolution model.

[0148] The Fisher spatiotemporal projection model employs a projection approach, projecting multidimensional variables onto a one-dimensional surface for processing. This ensures that the projected vector distances between different classes of data minimize intra-class distances and maximize inter-class distances, thereby distinguishing samples from different classes in a one-dimensional space. Using the Fisher spatiotemporal projection model as the primary model, it can simultaneously process fixed infrared signals from multiple channels and fully extract the linear features of these signals in both spatial and temporal dimensions. This allows the output to serve as a basis for determining the type of wearing or removing the device.

[0149] The spatiotemporal convolution model adds temporal convolution to spatial convolution, forming a convolutional structure that can process both spatial and temporal information simultaneously. The model can accurately estimate pose or signal trajectory based on the spatial distribution of data and its changes over time. Using the spatiotemporal convolution model as a secondary model allows for the simultaneous processing of fixed infrared signals from multiple channels, fully extracting the nonlinear characteristics of the fixed infrared signals in both spatial and temporal dimensions. This enables the output to serve as a basis for determining the type of donning or removing action.

[0150] In this embodiment, the Fisher spatiotemporal projection model is used as the first model and the spatiotemporal convolution model is used as the second model. This fully utilizes the characteristics of the Fisher spatiotemporal projection model and the spatiotemporal convolution model to identify linear and nonlinear features of multi-dimensional data, i.e., multi-channel fixed infrared signals. Furthermore, it classifies the types of donning and putting on actions represented by the signal performance of the fixed infrared signals. This allows both the first and second output results to be used as the basis for judging the type of donning and putting on actions, thereby improving the convenience, stability, and accuracy of determining donning and putting on action information.

[0151] In some embodiments, the sample data used to train the model for determining the wearing and removing action information is obtained as follows: Infrared signal segments corresponding to each type of wearing and removing action are extracted from infrared sample signals containing multiple wearing and removing actions. Each wearing action corresponds to multiple infrared signal segments, and for the same wearing action, the position of the action marker time in different infrared signal segments is different. Similarly, each removing action corresponds to multiple infrared signal segments, and for the same removing action, the position of the action marker time in different infrared signal segments is different.

[0152] The model can be determined by training on sample data to obtain information about the wearing and removing actions. When acquiring sample data, infrared signal segments corresponding to each type of wearing and removing action are extracted from infrared sample signals containing multiple wearing and removing actions. The infrared sample signals and the fixed infrared signals are acquired using the same infrared sensor, and the infrared sample signals are historical signals acquired during the model training phase. If the fixed infrared signal used to determine the wearing and removing action information is multi-channel, then the infrared sample signals are also multi-channel, with the same number of channels as the fixed infrared signal.

[0153] Since the infrared sample signal includes multiple wearing and removing actions, by extracting signals from the infrared sample signal, multiple infrared signal segments corresponding to each type of wearing and removing action can be obtained. The infrared signal segments corresponding to each type of wearing and removing action can be used as sample data to train the model for determining wearing and removing action information.

[0154] Each wearing or removing action corresponds to multiple infrared signal segments. For the same wearing or removing action, the position of the action marker time in different infrared signal segments is different. That is, there is a partial overlap between the multiple infrared signal segments corresponding to the same wearing or removing action, and the action marker time corresponding to the wearing or removing action is located within the overlapping part. This allows each wearing or removing action to correspond to multiple infrared signal segments, increasing the number of sample data and thus improving the robustness and sample distribution of the training model for determining wearing and removing action information.

[0155] For example, for any wearing action, the time of the abnormal peak point in the infrared sample signal corresponding to the wearing action can be used as the action marker time. The center time of a window with a window length of a first preset time period is set at the action marker time, and the infrared sample signal within the window is extracted as an infrared signal segment corresponding to the wearing action. Then, the window is slid forward and backward in steps of, for example, 0.5 seconds, and the infrared sample signal within the window at each sliding position is extracted as multiple other infrared signal segments corresponding to the wearing action, until the action marker time moves to the edge of the window and the window sliding is stopped, thus obtaining multiple infrared signal segments corresponding to the wearing action.

[0156] In this embodiment, infrared signal segments corresponding to each type of wearing and removing action are extracted from the infrared sample signals containing multiple wearing and removing actions. This enables the acquisition of sample data for training the model for determining wearing and removing action information, and ensures that each wearing or removing action corresponds to multiple infrared signal segments, thereby increasing the number of sample data and improving the robustness and sample distribution of the training model for determining wearing and removing action information.

[0157] After obtaining the sample data, the Fisher spatiotemporal projection model and the spatiotemporal convolution model can be trained respectively based on the sample data to obtain the removal and wearing action information determination model as the first model and the second model. For example, the sample data includes multiple infrared signal segments corresponding to each removal and wearing action type in 4 channels. The duration of each infrared signal segment is, for example, 3 seconds, and the number of sampling points in each infrared signal segment is 75 at a sampling frequency of 25 Hz.

[0158] For Fisher's spatiotemporal projection model, such as Figure 12 As shown, during the training phase, spatial and temporal projectors are learned based on the Fisher judgment criterion. The data dimension is (4, 75), meaning the sample data of each infrared signal band with 4 signal channels and 75 sampling points is compressed to a dimension of (1, 15) by the spatial projector, and then compressed to a dimension of (1, 4) by the temporal projector. Features in this dimension are then used to create classifiers such as Support Vector Machines (SVM) or tree-based algorithms (XGB). The spatial projector divides the sample data into 15 time points in the temporal dimension, using 5 sampling points as a small window. At each time point, the optimal projection filter for compressing the 4 channels of the signal at that time point is calculated based on the sample data. This compresses the data from 4 channels into 1 channel, ensuring that the compressed data maximizes the classification distance of the class samples. Afterwards, the compressed data still has 15 points in the temporal dimension that can represent temporal information. The projection filter is further trained in the temporal dimension to maximize the class discriminative power of the compressed temporal features.

[0159] Fisher's judgment criterion can be expressed, for example, by the following formula (1):

[0160]

[0161] Make J F The largest weight vector W is transformed into a derivative problem of formula (2) as shown below using the Lagrange multiplication operator:

[0162] L(W,γ)=W T S b W-γ(W T S w Wc)(2)

[0163] Wherein, S in formula (1) b Let S be the inter-class scatter matrix. w Within-class scatter matrix, W is the projection vector, S b and S w For example, it can be obtained by calculating the mean and covariance of the sample data. Furthermore, the above criteria can be applied to... Figure 12The spatial projection vector W is obtained in the spatial and temporal dimensions shown. s and time projection vector W t The final discriminant expression is obtained, which can be, for example, the following formula (3):

[0164]

[0165] Where k represents the spatial filter under the k-th window, n represents the n-th sampling point, and y n This represents the projected feature vector, which will ultimately be y n The data is fed into a classifier to obtain classification probabilities, enabling the trained Fisher spatiotemporal projection model to fully extract the discriminative features of the sample data in both spatial and temporal dimensions.

[0166] For spatiotemporal convolutional models, such as Figure 13 As shown, firstly, sample data of dimension (4, 75) is fed into the network through two conventional convolutional layers. The kernel size of the conventional convolution can be set to (1, c1), and the number of kernels can be set to 3. The size of c1 can be adjusted according to the desired training effect. Padding strategies such as padding same are used to ensure the dimension of the convolutional feature map matches the input. Next, the feature map is passed through a depthwise separable convolutional layer, compressing the 4 channels of data into 1 channel. The kernel size of the depthwise separable convolution can be set to (4, 1). Then, the feature map is passed through a separable convolutional layer, with a kernel size of (1, c2) and the number of kernels set to 3. The size of c2 can be adjusted according to the desired training effect. This layer compresses the temporal dimension features and expands the feature dimension. Finally, the feature map is passed through a flattening layer to convert it into a 1-dimensional feature vector, and then through a fully connected layer and an activation function to output the three-class classification result. A batch normalization (BN) layer can be added before each convolutional step for feature normalization. The network's activation function can be set to, for example, the softmax function, and an optimizer such as Adam can be used.

[0167] In some embodiments, reference Figure 14 As shown, after performing liveness detection in response to whether the wearing action is a wearing action or a removal action, the wearable device control method further includes:

[0168] S310. When the type of action of putting on or taking off is a wearing action, in response to the detection result of the liveness detection, the wearing state of the wearable device is determined to be the wrist state.

[0169] In step S310, when the type of wearing action is wearing action and triggers liveness detection, if the detection result of liveness detection is live, it means that a live target was detected after the wearing action occurred. At this time, it is considered that the user has worn the wearable device on the wrist, and the wearing status of the wearable device can be determined as the wrist state, and physiological indicators such as blood oxygen and heart rate can be detected.

[0170] S320. When the type of wearing action is removal, in response to the detection result of liveness detection being non-live, the wearing state of the wearable device is determined to be wrist-free.

[0171] In step S320, when the type of wearing action is removal action and a liveness detection is triggered, if the detection result of liveness detection is non-live, it means that no live target was detected after the removal action occurred. At this time, it is considered that the user has removed the wearable device from the wrist, the wearing status of the wearable device can be determined as wrist-free state, and the detection of physiological indicators such as blood oxygen and heart rate can be turned off.

[0172] In this embodiment, when a wearing action triggers liveness detection, if the liveness detection result indicates a live person, the wearable device is determined to be in a wrist-worn state. Conversely, when a removal action triggers liveness detection, if the liveness detection result indicates a non-live person, the wearable device is determined to be in a wrist-free state. This achieves the determination of the wearable device's wearing state. By using both the wearing action type and the liveness detection result as the basis for determining the wearing state, the accuracy of determining the wearing state is improved, ensuring that the wearing state simultaneously satisfies both the corresponding wearing action type and the liveness detection result.

[0173] In one exemplary embodiment, a wearable device control method is provided, applied to a wearable device, with reference to... Figure 15 As shown, the wearable device control method includes:

[0174] S1. Continuously acquire acceleration signals and multiple fixed infrared signals;

[0175] S2. Filter the acceleration signal within the second preset time period before the current time to obtain the filtered signal;

[0176] S3. Based on the filtered signal, determine the action level information, which is used to characterize the action level within the second preset time period before the current moment.

[0177] S4. Calculate the main frequency information of the acceleration signal within multiple fourth preset time periods. The termination times of the multiple fourth preset time periods are set at intervals within the third preset time period after the current time. The fourth preset time period is longer than the third preset time period.

[0178] S5. Based on the main frequency information, determine the action cycle information. The action cycle information is used to characterize whether the action within the third preset time period after the current moment is periodic.

[0179] S6. In response to an action level greater than a preset action level and an action that is not periodic, acquire multiple fixed infrared signals within a first preset time period before the current moment.

[0180] S7. Perform preprocessing on each fixed infrared signal to obtain a preprocessed signal that corresponds to each fixed infrared signal.

[0181] S8. Input multiple preprocessed signals into the Fisher spatiotemporal projection model and the spatiotemporal convolution model respectively to obtain the first output result of the Fisher spatiotemporal projection model and the second output result of the spatiotemporal convolution model.

[0182] S9. Based on the ambient light signal, determine the ambient light change information. The ambient light change information is used to characterize the degree of change of ambient light in the first preset time period before the current moment.

[0183] S10. Based on the first output of the Fisher spatiotemporal projection model, the second output of the spatiotemporal convolution model, and ambient light change information, determine the donning and removing action information;

[0184] S11. In response to the action type of putting on / removing being a wearing action, perform liveness detection;

[0185] S12. In response to the detection result of the liveness detection being live, determine that the wearable device is in the wrist state.

[0186] S13. In response to the action type of removing or putting on being a removal action, perform liveness detection;

[0187] S14. In response to the detection result of liveness detection being non-live, determine that the wearable device is in the wrist-free state.

[0188] In this embodiment, by determining the wearing and removing action information and performing liveness detection when the wearing and removing action type represented by the information is either wearing or removing, adaptive triggering of liveness detection is achieved, providing a basis for determining the wearing status. Using wearing and removing actions as trigger conditions for liveness detection ensures that liveness detection only occurs when the wearable device is worn or removed, guaranteeing that the triggering of liveness detection is compatible with detection requirements, reducing the detection trigger rate and interference with detection results, and improving the user experience.

[0189] In one exemplary embodiment, a wearable device control device is provided, applied to a wearable device, with reference to... Figure 16As shown, the wearable device control device includes a determination module 10 and a detection module 20. The determination module 10 is used to determine the wearing / removal action information, which characterizes the type of wearing / removal action, including wearing action, removal action, and no wearing / removal action. The detection module 20 is used to perform liveness detection in response to whether the wearing / removal action is a wearing action or a removal action.

[0190] In this embodiment, the determination module 10 determines the wearing / removing action information, and performs liveness detection when the wearing / removing action type represented by the action information is either a wearing action or a removal action. This achieves adaptive triggering of liveness detection, providing a basis for determining the wearing status. By using wearing and removal actions as triggering conditions for liveness detection, liveness detection is performed only when the wearable device is worn or removed, ensuring that the triggering of liveness detection is compatible with detection requirements, reducing the detection trigger rate and interference with detection results, and improving the user experience.

[0191] In one embodiment, the determining module 10 is further configured to: acquire a fixed infrared signal within a first preset time period before the current time; and determine the putting on / taking off action information based on the fixed infrared signal.

[0192] In one embodiment, the determining module 10 is further configured to: continuously acquire acceleration signals and fixed infrared signals; determine action level information based on the acceleration signals, wherein the action level information is used to characterize the action level within a second preset time period before the current moment; and acquire fixed infrared signals within a first preset time period before the current moment in response to an action level greater than a preset action level.

[0193] In one embodiment, the determining module 10 is further configured to: filter the acceleration signal within a second preset time period before the current time to obtain a filtered signal; and determine the action level information based on the filtered signal.

[0194] In one embodiment, the determining module 10 is further configured to: determine action cycle information based on the acceleration signal, wherein the action cycle information is used to characterize whether the action within a third preset time period after the current moment is periodic; and in response to an action level greater than a preset action level and the action not being periodic, acquire a fixed infrared signal within a first preset time period before the current moment.

[0195] In one embodiment, the determining module 10 is further configured to: calculate the main frequency information of the acceleration signal within a plurality of fourth preset time periods, wherein the termination times of the plurality of fourth preset time periods are set at intervals within a third preset time period after the current time, and the fourth preset time period is longer than the third preset time period; and determine the action cycle information based on each main frequency information.

[0196] In one embodiment, the determining module 10 is further configured to: preprocess the fixed infrared signal to obtain a preprocessed signal; input the preprocessed signal to the removal and wearing action information determining model; and determine the removal and wearing action information based on the output result of the removal and wearing action information determining model.

[0197] In one embodiment, there are multiple fixed infrared signals, and the model for determining the wearing and removing action information includes a first model and a second model. The first model is capable of linear feature recognition, and the second model is capable of nonlinear feature recognition. The determining module 10 is further configured to: preprocess each fixed infrared signal to obtain a preprocessed signal corresponding to each fixed infrared signal; input the multiple preprocessed signals to the first model and the second model respectively to obtain a first output result of the first model and a second output result of the second model; and perform a fusion decision on the first output result and the second output result to determine the wearing and removing action information.

[0198] In one embodiment, the determining module 10 is further configured to: determine ambient light change information based on the ambient light signal, wherein the ambient light change information is used to characterize the degree of change of ambient light within a first preset time period before the current moment; and determine the removal and wearing action information based on the output result and the ambient light change information.

[0199] In one embodiment, the first model includes a Fisher spatiotemporal projection model; and / or, the second model includes a spatiotemporal convolution model.

[0200] In one embodiment, the sample data used to train the model for determining the information of wearing and removing actions is obtained as follows: infrared signal segments corresponding to each type of wearing and removing action are extracted from infrared sample signals containing multiple wearing and removing actions to obtain sample data; wherein, the action marker time corresponding to each wearing action corresponds to multiple infrared signal segments, and for the same wearing action, the position of the action marker time in different infrared signal segments is different; the action marker time corresponding to each removing action corresponds to multiple infrared signal segments, and for the same removing action, the position of the action marker time in different infrared signal segments is different.

[0201] In one embodiment, after performing liveness detection in response to whether the action type is wearing or removing, the detection module 20 is further configured to: determine the wearing state of the wearable device as wrist-worn if the detection result of the liveness detection is live when the action type is wearing; and determine the wearing state of the wearable device as wrist-free if the detection result of the liveness detection is not live when the action type is removing.

[0202] In one exemplary embodiment, an electronic device is provided, which may include, for example, a wearable device such as a watch or a bracelet that can be worn on a specific part of a user's body and has physiological indicator detection functions.

[0203] refer to Figure 17 As shown, the electronic device may include one or more of the following components: processing component 101, memory 102, power component 103, multimedia component 104, audio component 105, input / output (I / O) interface 106, sensor component 107, and communication component 108.

[0204] Processing component 101 typically controls the overall operation of an electronic device, such as operations associated with display, telephone calls, data communication, camera operation, and recording. Processing component 101 may include one or more processors 109 to execute instructions to perform all or part of the steps of the methods described above. Furthermore, processing component 101 may include one or more modules to facilitate interaction between processing component 101 and other components. For example, processing component 101 may include a multimedia module to facilitate interaction between multimedia component 104 and processing component 101.

[0205] Memory 102 is configured to store various types of data to support the operation of the electronic device. Examples of such data include instructions for any application or method used to operate on the electronic device, contact data, phonebook data, messages, pictures, videos, etc. Memory 102 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0206] Power component 103 provides power to various components of the electronic device. Power component 103 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the electronic device.

[0207] Multimedia component 104 includes a screen that provides an output interface between the electronic device and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of touch or swipe actions but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 104 includes a front-facing camera and / or a rear-facing camera. When the electronic device is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0208] Audio component 105 is configured to output and / or input audio signals. For example, audio component 105 includes a microphone (MIC) configured to receive external audio signals when the electronic device is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 102 or transmitted via communication component 108. In some embodiments, audio component 105 also includes a speaker for outputting audio signals.

[0209] I / O interface 106 provides an interface between processing component 101 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0210] Sensor assembly 107 includes one or more sensors for providing state assessments of various aspects of the electronic device. For example, sensor assembly 107 can detect the on / off state of the electronic device, the relative positioning of components such as the display and keypad of the electronic device, changes in the position of the electronic device or a component of the electronic device, the presence or absence of user contact with the electronic device, the orientation or acceleration / deceleration of the electronic device, and temperature changes of the electronic device. Sensor assembly 107 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 107 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 107 may also include an accelerometer, a gyroscope, a magnetometer, a pressure sensor, or a temperature sensor.

[0211] Communication component 108 is configured to facilitate wired or wireless communication between electronic devices and other devices. Devices can access wireless networks based on communication standards, such as WiFi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, communication component 108 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 108 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0212] In an exemplary embodiment, the electronic device may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the wearable device control method applied to the electronic device described above.

[0213] In one exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 102 including instructions, which can be executed by a processor 109 of an electronic device to perform the wearable device control method applied to the electronic device described above. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc. When the instructions in the storage medium are executed by the processor 109 of the electronic device, the electronic device is able to perform the wearable device control method shown in the above embodiments.

[0214] In one exemplary embodiment, a computer program product is also provided, including a computer program that, when executed by processor 109, implements the wearable device control method shown in the above embodiments.

[0215] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0216] It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A wearable device control method, characterized in that, The wearable device control method includes: Determine the information on wearing and removing actions, which is used to characterize the type of wearing and removing actions, including wearing actions, removing actions, and no wearing / removing actions; In response to the action type being either wearing or removing, a liveness detection is performed.

2. The wearable device control method according to claim 1, characterized in that, The information for determining the wearing and removing actions includes: Acquire the fixed infrared signal within the first preset time period before the current time; The information regarding the putting on / taking off action is determined based on the fixed infrared signal.

3. The wearable device control method according to claim 2, characterized in that, The acquisition of the fixed infrared signal within the first preset time period prior to the current moment includes: Continuously acquire acceleration signals and the fixed infrared signals; Based on the acceleration signal, action level information is determined, which is used to characterize the action level within a second preset time period before the current moment. In response to the action level being greater than a preset action level, a fixed infrared signal within the first preset time period prior to the current moment is acquired.

4. The wearable device control method according to claim 3, characterized in that, The determination of action level information based on the acceleration signal includes: The acceleration signal within the second preset time period before the current moment is filtered to obtain a filtered signal; The action level information is determined based on the filtered signal.

5. The wearable device control method according to claim 3, characterized in that, The method of acquiring the fixed infrared signal within a first preset time period prior to the current moment further includes: Based on the acceleration signal, action cycle information is determined, which is used to characterize whether the action within a third preset time period after the current moment is periodic. The step of responding to an action level greater than a preset action level by acquiring a fixed infrared signal within the first preset time period prior to the current moment includes: In response to the action level being greater than a preset action level and the action not being periodic, a fixed infrared signal within the first preset time period prior to the current moment is acquired.

6. The wearable device control method according to claim 5, characterized in that, Determining the motion cycle information based on the acceleration signal includes: Calculate the main frequency information of the acceleration signal within multiple fourth preset time periods. The termination times of the multiple fourth preset time periods are set at intervals within a third preset time period after the current time. The fourth preset time periods are longer than the third preset time periods. Based on the aforementioned main frequency information, the action cycle information is determined.

7. The wearable device control method according to claim 2, characterized in that, The process of determining the donning / removing action information based on the fixed infrared signal includes: The fixed infrared signal is preprocessed to obtain a preprocessed signal; The preprocessed signal is input into the donning and wearing action information determination model, and the donning and wearing action information is determined based on the output of the donning and wearing action information determination model.

8. The wearable device control method according to claim 7, characterized in that, The fixed infrared signals are multiple, and the model for determining the wearing and removing action information includes a first model and a second model. The first model is capable of linear feature recognition, and the second model is capable of nonlinear feature recognition. The preprocessing of the fixed infrared signal to obtain a preprocessed signal includes: Each of the fixed infrared signals is preprocessed to obtain a preprocessed signal that corresponds one-to-one with each of the fixed infrared signals. The step of inputting the preprocessed signal into the donning / removal action information determination model, and determining the donning / removal action information based on the output of the donning / removal action information determination model, includes: The preprocessed signals are respectively input into the first model and the second model to obtain the first output result of the first model and the second output result of the second model; The first output result and the second output result are fused together to determine the wearing and removing action information.

9. The wearable device control method according to claim 7, characterized in that, The wearable device control method further includes: Based on the ambient light signal, ambient light change information is determined, which is used to characterize the degree of change of ambient light within a first preset time period before the current moment. The step of determining the wearing / removing action information based on the output of the model includes: determining the wearing / removing action information based on the output and the ambient light change information.

10. The wearable device control method according to claim 8, characterized in that, The first model includes the Fisher spatiotemporal projection model; and / or, The second model includes a spatiotemporal convolution model.

11. The wearable device control method according to claim 7, characterized in that, The sample data used to train the model for determining the information of the donning and putting on actions is obtained in the following way: The sample data is obtained by extracting infrared signal segments corresponding to each type of wearing and removing action from infrared sample signals that contain multiple wearing and removing actions. Each of the wearing actions corresponds to multiple infrared signal segments, and for the same wearing action, the position of the action mark time in different infrared signal segments is different. Each removal action corresponds to a mark time in multiple infrared signal segments. For the same removal action, the mark time is located in different infrared signal segments.

12. The wearable device control method according to any one of claims 1 to 11, characterized in that, After performing liveness detection in response to the action being either wearing or removing the device, the wearable device control method further includes: When the type of the wearing and removing action is a wearing action, in response to the detection result of the liveness detection being a liveness, the wearing state of the wearable device is determined to be a wrist state; When the type of wearing action is removal, in response to the detection result of the liveness detection being non-live, the wearing state of the wearable device is determined to be wrist-free.

13. A wearable device control device, characterized in that, The wearable device control device includes: A determining module is used to determine the wearing / removing action information, which is used to characterize the wearing / removing action type, including wearing action, removal action, and no wearing / removing action. The detection module is used to perform liveness detection in response to whether the wearing or removing action is a wearing action or a removing action.

14. An electronic device, characterized in that, The electronic device includes: processor; Memory used to store processor-executable instructions; The processor is configured to perform the wearable device control method as described in any one of claims 1 to 12.

15. A non-transitory computer-readable storage medium, characterized in that, When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device is able to perform the wearable device control method as described in any one of claims 1 to 12.

16. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the wearable device control method as described in any one of claims 1 to 12.

Images