WO2023240512A1 - 跌倒检测方法、装置、耳机及存储介质 - Google Patents

跌倒检测方法、装置、耳机及存储介质 Download PDF

Info

Publication number
WO2023240512A1
WO2023240512A1 PCT/CN2022/099024 CN2022099024W WO2023240512A1 WO 2023240512 A1 WO2023240512 A1 WO 2023240512A1 CN 2022099024 W CN2022099024 W CN 2022099024W WO 2023240512 A1 WO2023240512 A1 WO 2023240512A1
Authority
WO
WIPO (PCT)
Prior art keywords
audio
ear canal
user
audio signal
target
Prior art date
Application number
PCT/CN2022/099024
Other languages
English (en)
French (fr)
Inventor
周岭松
Original Assignee
北京小米移动软件有限公司
北京小米松果电子有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京小米移动软件有限公司, 北京小米松果电子有限公司 filed Critical 北京小米移动软件有限公司
Priority to CN202280004515.1A priority Critical patent/CN117597065A/zh
Priority to PCT/CN2022/099024 priority patent/WO2023240512A1/zh
Publication of WO2023240512A1 publication Critical patent/WO2023240512A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb

Definitions

  • the present disclosure relates to the field of information processing technology but is not limited to the field of information processing technology, and in particular, to a fall detection method, device, earphones and storage media.
  • the health status of the detected object can be detected through health monitoring equipment. For example, to detect the fall of the target group, by wearing corresponding sensors on the target group's body, or letting the target group wear clothing with corresponding sensors, etc., it is determined whether the target group has fallen based on the signal detected by the sensor.
  • Target groups can include the elderly and people with limited mobility.
  • Embodiments of the present disclosure provide a fall detection method, device, earphones and storage media.
  • a first aspect of an embodiment of the present disclosure provides a fall detection method, applied to earphones, where the earphones include a feedback microphone, and the method includes:
  • the feedback microphone collects the audio signal in the ear canal to obtain the ear canal audio signal; wherein the ear canal audio signal includes: when the earphone is worn by the user, the user's body collides with the ground when he falls. The generated vibration is transmitted to the ear canal through bone conduction to produce an audio signal;
  • the features to be identified are input into the fall detection model to obtain detection results; wherein the detection results are used to indicate that the user has fallen.
  • a second aspect of the embodiment of the present disclosure provides a fall detection device, applied to headphones, where the headphones include a feedback microphone, and the device includes:
  • the ear canal audio signal detection module is configured to collect the audio signal in the ear canal through the feedback microphone to obtain the ear canal audio signal; wherein the ear canal audio signal includes: when the earphone is worn by the user, When the user falls, the vibration generated by the collision between the body and the ground is transmitted to the ear canal through bone conduction to generate an audio signal;
  • An audio signal characteristic parameter acquisition module is configured to perform feature extraction on the ear canal audio signal to obtain audio signal characteristic parameters
  • the periodic information determination module is configured to filter the ear canal audio signal according to a preset frequency range to obtain a periodic signal of the preset frequency range; wherein the periodic signal includes a peak value or a valley value of the waveform;
  • a feature generation module to be identified configured to generate features to be identified according to the audio signal feature parameters and the number of peaks or valleys;
  • the detection module is configured to input the feature to be identified into the fall detection model to obtain a detection result; wherein the detection result is used to indicate that the user has fallen.
  • a third aspect of the present disclosure provides an earphone.
  • the earphone includes a housing and a controller, a feedback microphone, a feedforward microphone and a speaker provided on the housing; the feedforward microphone is connected to the controller. , used to collect audio data outside the ear canal and send it to the controller; the feedback microphone is connected to the controller, used to collect audio data inside the ear canal and send it to the controller; the controller includes a memory and a processor
  • the processor has executable computer instructions stored in the memory, and the processor can call the computer instructions stored in the memory to execute the method described in any of the above embodiments.
  • a fourth aspect of the embodiments of the present disclosure provides a computer storage medium that stores an executable program; after the executable program is executed by a processor, the fall detection method provided by the first aspect can be implemented.
  • the fall detection method provided by the embodiments of the present disclosure can be applied to earphones. Whether the user has fallen can be determined through the earphones without the need for other detection sensors, thereby improving the convenience of detecting the user's fall and improving the user experience.
  • Figure 1 is a schematic diagram of an earphone according to an exemplary embodiment
  • Figure 2 is a schematic diagram of an earphone according to an exemplary embodiment
  • Figure 3 is a schematic diagram of an earphone in a state of being worn by a user according to an exemplary embodiment
  • Figure 4 is a schematic diagram of another fall detection method according to an exemplary embodiment
  • Figure 5 is a schematic diagram of a periodic signal according to an exemplary embodiment
  • Figure 6 is a schematic diagram of a fall detection device according to an exemplary embodiment
  • FIG. 7 is a schematic structural diagram of an electronic device according to an exemplary embodiment.
  • first, second, third, etc. may be used to describe various information in the embodiments of the present disclosure, the information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other.
  • first information may also be called second information, and similarly, the second information may also be called first information.
  • word “if” as used herein may be interpreted as "when” or "when” or "in response to determining.”
  • fall detection for the elderly or people with mobility impairments is done through multiple sensors.
  • multiple acceleration sensors are deployed on the bodies of these people, or these people wear clothing equipped with acceleration sensors, etc., and it is determined whether the relevant user has fallen based on the change information of the signal detected by the acceleration sensor.
  • FIG. 1 a schematic diagram of a fall detection method provided by an embodiment of the present disclosure is shown.
  • the method can be applied to at least earphones, and the earphones can at least include a feedback microphone.
  • the method includes:
  • Step S100 collect the audio signal in the ear canal through the feedback microphone to obtain the ear canal audio signal; wherein, the ear canal audio signal includes: when the earphone is worn by the user, the vibration generated by the collision between the user's body and the ground when the user falls passes through the bone.
  • the audio signal produced by conduction is transmitted to the ear canal.
  • Step S200 perform feature extraction on the ear canal audio signal to obtain audio signal feature parameters.
  • Step S300 Filter the ear canal audio signal according to a preset frequency range to obtain a periodic signal in the preset frequency range; wherein the periodic signal includes peaks or valleys of the waveform.
  • Step S400 Generate features to be identified based on the audio signal feature parameters and the number of peaks or valleys.
  • Step S500 Input the features to be identified into the fall detection model to obtain detection results; where the detection results are at least used to indicate that the user has fallen.
  • Earphones can be of different shapes, including in-ear, semi-in-ear, and head-mounted earphones.
  • the communication method of the headset may include wired headsets and wireless headsets, and the wireless headsets may include Bluetooth headsets, such as true wireless stereo headsets (True Wireless Stereo, TWS). Headphones may also include devices such as hearing aids that have feedback microphones and are capable of implementing this scheme.
  • the feedback microphone in the headset can be located near the sound channel of the headset.
  • the feedback microphone is located in the ear canal and can collect audio signals in the ear canal, such as in-ear headphones.
  • the earphones are other types of earphones, such as semi-in-ear earphones and headphones, the feedback microphone can collect audio signals in the ear canal when the earphones are worn.
  • FIG. 2 a schematic diagram of an earphone is shown, including a feedback microphone A.
  • the feedback microphone A When the earphone is worn, the feedback microphone A may be located in the ear canal.
  • the headset may also include a feedforward microphone B.
  • the feedforward microphone B may be located on the stem of the headset.
  • the feedforward microphone When the headset is worn, the feedforward microphone is located outside the ear canal and can collect environmental audio signals from the external environment.
  • the headset may also include a call microphone C to collect audio signals sent by the user during a call.
  • the signal-to-noise ratio of the audio signal collected by feedback microphone A is higher than that of feedforward microphone B, so an ear canal audio signal with less noise and higher quality is collected through feedback microphone A.
  • the feedback microphone A can be located outside the ear canal or toward the ear canal, and can collect ear canal audio signals.
  • FIG. 3 is a schematic diagram of an earphone when being worn by a user
  • the earphone 1 blocks the ear canal 2 to a certain extent, forming an ear-blocking effect.
  • the ear canal audio signals collected by the feedback microphone include: when the headset is worn by the user, the vibration generated by the user during breathing is transmitted to the ear canal through bone conduction, that is, the audio signal 3.
  • step S100 when the earphones are in the wearing state, the ear canal will be blocked to a certain extent, resulting in a certain degree of ear blocking effect.
  • the reason for this is that some sounds will be conducted to the inner ear through the human bones. For example, due to the contact between the feet and the inner ear when walking. The ground contact generates vibrations, which are transmitted to the audio signals in the ear canal through bone conduction.
  • part of the sound from the bone conduction diffuses outward through the outer ear.
  • the ear canal is blocked to a certain extent, which reduces the sound from the bone conduction to the outside through the ear canal.
  • the amount of diffusion creates a certain degree of ear-blocking effect, also known as the occlusion effect.
  • the sound characteristics produced by the occlusion effect are characterized by strengthening low-frequency signals and weakening high-frequency signals.
  • the earphones block the ear canal to a certain extent, they form varying degrees of ear blocking effect. After the ear blocking effect occurs, the earphones block external audio signals from entering the ear canal, reducing the impact of external audio signals on the audio signals in the ear canal.
  • the feedback microphone can collect the audio signal in the ear canal to obtain the respiratory audio signal.
  • the respiratory audio signal includes the vibration generated by the contact between the user's feet and the ground when the earphone is worn by the user when walking, which is transmitted to the ear canal through bone conduction. The audio signal produced.
  • the vibration generated by the user's foot contact with the ground when walking can be transmitted to the ear canal through bone conduction, generating an audio signal.
  • the ear-blocking effect can amplify the audio signal, which facilitates the feedback microphone to collect the ear canal.
  • the audio signal in the canal is obtained to obtain the ear canal audio signal.
  • step S200 after obtaining the ear canal audio signal, feature extraction can be performed on the ear canal audio signal to obtain audio signal characteristic parameters.
  • Feature extraction methods can include a variety of methods, such as extracting corresponding features through feature extraction algorithms.
  • the obtained audio signal feature parameters can be Mel spectrum coefficients and Mel cepstrum coefficients (MFCC). Mel spectrum coefficients and Mel cepstral coefficients
  • MFCC Mel cepstrum coefficients
  • the numbers can all be 40-dimensional feature parameters. Of course, it can also be other characteristics of the ear canal audio signal.
  • the corresponding audio signal characteristics will also be different, and different ear canal audio signals have their own audio signal characteristics.
  • the audio signal characteristics of the audio signal generated when the user's body contacts the ground are transmitted to the ear canal through bone conduction when the user falls. This is different from the vibration generated when the user's body collides with the ground when the user does not fall. The vibration is transmitted through bone conduction.
  • the audio signal characteristics of the audio signal generated by the ear canal are also different from the audio signal characteristics of other audio signals generated by the vibration being transmitted to the ear canal through bone conduction, such as the audio signals of speech and external environment sound signals. feature.
  • the ear canal audio signal can be filtered according to the preset frequency range to obtain a periodic signal in the preset frequency range, and the periodic signal includes the peak value of the waveform.
  • the preset frequency range can be determined according to actual usage requirements or can be preset. It can also be determined based on the walking pace of a preset number of users. For example, the preset frequency range may be 1 Hz to 50 Hz, and the ear canal audio signal is low-pass filtered according to the preset frequency range.
  • the audio signals outside the preset frequency range in the ear canal audio signal are filtered out to obtain audio signals with frequencies within the preset frequency range.
  • the audio signals within the preset frequency range can be periodic signals.
  • the periodic signal can be a signal in the time domain.
  • the periodic signal is expressed in the form of a waveform, including the relationship between time and amplitude.
  • the peak value of the waveform in the waveform can be determined.
  • the peak of the waveform in the waveform indicates that the user's foot is in contact with the ground at the time corresponding to the peak, and one peak indicates that the foot is in contact with the ground once.
  • the number of steps taken by the user can be determined based on the number of peaks.
  • the troughs in the waveform can be determined based on the periodic signal.
  • the trough value in the waveform indicates that the user's foot is in contact with the ground at the time corresponding to the trough value, and one valley value indicates that the foot is in contact with the ground once.
  • the number of steps taken by the user can be determined based on the number of valleys.
  • Step S200 may be executed first, or step S300 may be executed first.
  • the features to be identified can be generated based on the number of peaks and the audio signal feature parameters, or the features to be identified can be generated based on the number of valleys and the audio signal feature parameters.
  • feature For example, the number of peaks and the audio signal characteristic parameters can be used together as a feature to be identified.
  • the feature to be identified includes information in two dimensions: the number of peaks and the audio signal characteristic parameters, and is used to determine the detection result.
  • the number of valley values and the audio signal characteristic parameters can also be used together as a feature to be identified.
  • the feature to be identified includes information in two dimensions: the number of valley values and the audio signal characteristic parameters, which are used to determine the detection result.
  • the obtained features to be identified are different.
  • the obtained features to be identified are different.
  • the feature to be identified changes.
  • the feature to be identified changes. This can reduce the situation where the features to be identified remain unchanged when the number of peaks and the characteristic parameters of the audio signal change at the same time, or when the number of valleys and the characteristic parameters of the audio signal change at the same time, thereby improving the accuracy of the detection results. .
  • step S500 the features to be identified are input into the fall detection model to obtain the detection results.
  • This detection result is at least used to indicate that the user has fallen.
  • the fall detection model is a detection model that has been trained in advance.
  • the vibration generated by the collision between the user's body and the ground when the user falls is transmitted to the ear canal through bone conduction.
  • the audio signal characteristics of the audio signal generated are different from the audio signal characteristics of other audio signals. , and the number of peaks or valleys in the corresponding periodic signals is also different, so inputting the features to be identified into the fall detection model can determine whether the user has fallen.
  • mapping table includes the mapping relationship between the features to be identified and the detection results. By looking up the mapping table, the detection results can be determined based on the features to be identified.
  • This disclosed example can obtain the user's ear canal audio signal through headphones, then process the ear canal audio signal, and use a fall detection model to detect whether the user has fallen.
  • the headset can determine whether the user has fallen. It reduces the difficulty and inconvenience of detecting a user's fall, improves the convenience of detecting a user's fall, reduces the discomfort and inconvenience caused to the user in the process of detecting a user's fall, and improves the user's experience.
  • the ear canal audio signal does not include that when the headset is worn by the user, the vibration generated by the collision between the user's body and the ground when the user falls is transmitted to the ear canal through bone conduction.
  • the audio signal is generated, through steps S200 to S500, a detection result indicating that the user has not fallen can be obtained, and the detection result can be output at this time.
  • FIG. 4 is a schematic diagram of another fall detection method.
  • the method also includes:
  • Step S10 Determine the target frame length of at least one frame period signal according to the duration corresponding to the preset number of steps.
  • Step S20 Determine the number of peaks or valleys included in each frame period signal within the target frame length; each step corresponds to a peak or valley.
  • the target frame length of at least one frame period signal can be determined based on the duration corresponding to the preset number of steps.
  • the preset number of steps can be determined according to actual application requirements, for example, it can be the duration of two steps, the duration of three steps, or the duration of four steps.
  • the duration corresponding to the preset number of steps can be in seconds, and the duration corresponding to the preset number of steps can be determined as the target frame length of the one-frame period signal, or the duration corresponding to the preset number of steps can be determined as the target frame length of the N-frame period signal.
  • N is equal to 1, that is, the duration corresponding to the preset number of steps is used to determine the target frame length of a frame period signal. This can improve the accuracy and convenience of determining the number of peaks or valleys.
  • the number of peaks or valleys included in each frame period signal can be determined within the target frame length. Since the periodic signal is expressed in the form of a waveform, the peak or trough of the waveform can also be determined based on the periodic signal. The peak value of the wave crest can be determined based on the wave crest, and the valley value of the wave trough can be determined based on the wave trough interface. According to the target frame length, the number of peaks or valleys within the target frame length is determined. Each step corresponds to a peak or valley value. Since the user's feet vibrate when they are in contact with the ground, the intensity of the audio signal is greater than the intensity of the corresponding audio signal when the feet are not in contact with the ground. Therefore, the peak or valley value within the target frame length is The number of can represent the number of steps taken by the user, that is, the number of times the user's feet are in contact with the ground.
  • This method can determine the number of corresponding peaks or valleys in each frame period signal.
  • the peak detection algorithm can be used to detect the signal of each frame period and determine the peak value in the signal of each frame period.
  • the number of valley values can be detected by the valley value detection algorithm on each frame period signal, and the valley value in each frame period signal can be determined.
  • the peak detection algorithm can detect the peaks of the waveform in the periodic signal, and then determine the corresponding peak value.
  • the valley detection algorithm can detect the waveform valleys in the periodic signal, and then determine the corresponding valley value.
  • features to be identified may be generated based on the number of peaks in the signal of one frame period, or features to be identified may be generated based on the number of peaks in the signal of multiple frame periods.
  • the features to be identified can be generated based on the average number of peaks in multiple consecutive periodic signals and the audio signal feature parameters.
  • each target frame length is the duration corresponding to the two steps, and the duration is 1 second.
  • each periodic signal includes three peaks within the target frame length. Taking the periodic signal corresponding to the leftmost target frame length as an example, each peak indicates that the foot is in contact with the ground, indicating that the user is passing through the foot.
  • the duration from the leftmost peak to the middle peak is the duration of the first step, and the duration from the middle peak to the rightmost peak is the duration of the next step.
  • the peak value of the corresponding wave peak or the valley value of the wave trough within each target frame length is represented by a dot, and the peak value or valley value represented by the dot can be determined through the corresponding detection algorithm.
  • the ear canal audio signal includes when the headset is worn by the user, and the vibration generated by the collision between the user's body and the ground when the user falls is transmitted to the ear canal through bone conduction.
  • the number of peaks or valleys included in each periodic signal of the target frame length is greater than the corresponding peaks or valleys when walking normally without falling. quantity.
  • the fall detection model is obtained by pre-training the initial neural network model based on the fall information training sample set using machine learning.
  • the structure of the initial neural network model is not limited. After training through the training sample set through machine learning, the recognition result can be output based on the features to be recognized.
  • the training sample set includes a positive sample set, and the positive sample set includes a plurality of positive samples.
  • Each positive sample includes: the audio features of the target collision, the first number and the first label.
  • the audio characteristics of the target collision are the audio characteristics obtained by collecting the audio signals in the ear canal through the feedback microphone when the user's body collides with the ground when the headset is worn by the user. After the feedback microphone in the earphone collects the low audio signal in the ear canal, the earphone can perform feature extraction on the collected audio signal to obtain the audio characteristics of the target collision.
  • the first quantity is: the number of peaks or valleys in the waveform included in the periodic signal of the preset frequency range obtained after filtering the audio characteristics of the target collision according to the preset frequency range.
  • the first label is used to represent the audio feature of the target collision and the first number corresponds to the output of the initial neural network model.
  • Each positive sample in the positive sample set is input into the initial neural network model, and the first label is used as the output of the initial neural network.
  • the initial neural network model is trained to obtain a fall detection model.
  • the number of positive samples in the positive sample set can be determined according to actual needs. The greater the number, the higher the detection accuracy of the trained fall detection model.
  • the audio features of the target collision may include Mel spectrum feature parameters and Mel cepstrum feature parameters, etc., and both the Mel spectrum coefficients and the Mel cepstrum coefficients may be 40-dimensional coefficients.
  • the audio features of the target collision included in different positive samples correspond to different postures and/or number of collisions between the user's body and the ground when he falls.
  • Each positive sample can be the audio feature of the audio signal of different parts of the body colliding with the ground when different users fall in different postures, and the audio features of the target collision included in the different positive samples correspond to the audio features of the user's body and the ground when they fall.
  • the posture of the collision, the location of the collision with the ground, and/or the number of collisions are different.
  • Each positive sample can also be the audio features of the audio signals of different parts of the body colliding with the ground when the same user falls in different postures, and the audio features of the target collision included in different positive samples correspond to the audio features of the user's body when he falls.
  • the posture of the collision with the ground, the location of the collision with the ground, and/or the number of collisions are different.
  • the audio features included in the positive sample 1 are the audio features of the collision between the body and the ground when user 1 falls in posture 1.
  • the number of collisions between different parts of the body and the ground when user 1 falls in posture 1 is the number of times 1, and the number of collisions with the ground is The parts border the hands and knees.
  • the audio features included in the positive sample 2 are the audio features of user 2's body colliding with the ground when he fell in posture 2.
  • the number of times that different parts of the body collided with the ground when user 2 fell in posture 2 is the number of times 2, and the parts that collided with the ground are Including hands and buttocks.
  • the audio features included in the positive sample 3 are the audio features of the collision between user 1's body and the ground when he fell in posture 2.
  • the number of times that different parts of the body collided with the ground when user 1 fell in posture 2 is the number 3, and the parts that collided with the ground are Includes back and head.
  • the first number is the number of peaks or valleys within the target frame length.
  • the corresponding first number in the positive sample and the number of peaks or valleys that generate the features to be identified are determined within the same frame length, which can reduce variables and improve detection accuracy.
  • the training sample set also includes a negative sample set, and the negative sample set includes multiple negative samples.
  • Each negative sample includes: audio features of non-target collisions, a second quantity, and a second label.
  • the audio characteristics of non-target collisions are the audio characteristics obtained by collecting the audio signals in the ear canal through the feedback microphone when the headset is worn by the user, except when the user's body collides with the ground when falling.
  • the second quantity is: the number of peaks or valleys of the waveform included in the periodic signal of the preset frequency range obtained after filtering the audio characteristics of the non-target collision according to the preset frequency range.
  • the second label is used to represent the audio features of the non-target collision and the output of the initial neural network model corresponding to the second quantity.
  • the audio features included in the negative samples are different from the audio features included in the positive samples.
  • the positive samples include the audio features of the audio signals collected by the feedback microphone when the body collides with the ground in various falling states.
  • the negative samples include the non-audio features collected by the feedback microphone. Audio characteristics of various audio signals in the ear canal during the fall state. In the non-fall state, the audio signal in the ear canal may include audio features of environmental audio, speaking audio, and audio generated by the user's other interactive operations. These audio features are different from the audio features of the target collision included in the positive sample.
  • Each negative sample in the negative sample set is input into the initial neural network model, and the second label is used as the output of the initial neural network.
  • the initial neural network model is trained to obtain a fall detection model.
  • the number of negative samples in the negative sample set can be determined according to actual needs. The greater the number, the higher the detection accuracy of the trained fall detection model.
  • the audio features of non-target collisions may include Mel spectrum feature parameters and Mel cepstrum feature parameters, etc., and both the Mel spectrum coefficients and the Mel cepstrum coefficients may be 40-dimensional coefficients.
  • the second number is the number of peaks or valleys within the target frame length.
  • the corresponding second number in the negative sample and the number of peaks or valleys that generate the features to be identified are both determined within the same frame length, which can reduce variables and improve detection accuracy.
  • the initial neural network model is trained using positive samples and negative samples to obtain a fall detection model, which improves the detection capability of the fall detection model and results in more accurate detection results.
  • the fall detection method further includes:
  • the preset device can be a mobile phone, a tablet, etc., or a device held by a user who has a social relationship with the detected user. For example, if the user is an elderly person, the default device can be the electronic device of the caregiver.
  • the prompt information can be a pop-up message, a sound prompt message or a short message, etc.
  • Figure 6 is a schematic diagram of a fall detection device, which is applied to a headset, the headset includes a feedback microphone, and the device includes:
  • the ear canal audio signal detection module 1 is configured to collect the audio signal in the ear canal through the feedback microphone to obtain the ear canal audio signal; wherein the ear canal audio signal includes: when the earphone is worn by the user , the audio signal generated by the vibration generated by the collision between the user's body and the ground when the user falls is transmitted to the ear canal through bone conduction;
  • the audio signal characteristic parameter acquisition module 2 is configured to perform feature extraction on the ear canal audio signal to obtain the audio signal characteristic parameters
  • the periodic signal determination module 3 is configured to filter the ear canal audio signal according to a preset frequency range to obtain a periodic signal of the preset frequency range; wherein the periodic signal includes the peak value or the valley value of the waveform;
  • the feature generation module 4 to be identified is configured to generate features to be identified according to the audio signal feature parameters and the number of peaks or valleys;
  • the detection module 5 is configured to input the features to be identified into the fall detection model to obtain a detection result; wherein the detection result is at least used to indicate that the user has fallen.
  • the device further includes:
  • a target frame length determination module configured to determine the target frame length of at least one frame of the periodic signal based on the duration corresponding to the preset number of steps
  • a quantity determination module configured to determine the number of the peak values or the valley values included in the periodic signal of each frame within the target frame length; wherein each step corresponds to one of the peak values or the valley values.
  • the fall detection model is obtained by pre-training the initial neural network model based on the fall information training sample set using machine learning.
  • the training sample set includes a positive sample set, and the positive sample set includes a plurality of positive samples;
  • Each of the positive samples includes: an audio feature of a target collision, a first quantity and a first label;
  • the audio characteristics of the target collision are the audio characteristics obtained by collecting audio signals in the ear canal through the feedback microphone when the user's body collides with the ground when the user falls and the headset is worn by the user;
  • the first number is: the number of peaks or valleys in the waveform included in the periodic signal of the preset frequency range obtained after filtering the audio characteristics of the target collision according to the preset frequency range;
  • the first label is used to represent the audio feature of the target collision and the first quantity corresponds to the output of the initial neural network model.
  • the first number is the number of peaks or valleys within the target frame length.
  • the audio features of the target collision included in the different positive samples correspond to different postures of the user's body colliding with the ground when he fell, parts of the body colliding with the ground, and/or the number of collisions.
  • the training sample set also includes a negative sample set, and the negative sample set includes a plurality of negative samples;
  • Each said negative sample includes: an audio feature of a non-target collision, a second quantity and a second label;
  • the audio characteristics of the non-target collision are the audio characteristics obtained by collecting the audio signals in the ear canal through the feedback microphone when the user is wearing the earphones and the user collides with the ground except when the user falls;
  • the second number is: the number of peaks or valleys in the waveform included in the periodic signal of the preset frequency range obtained after filtering the audio characteristics of the non-target collision according to the preset frequency range;
  • the second label is used to represent the audio feature of the non-target collision and the second quantity corresponds to the output of the initial neural network model.
  • the second number is the number of peaks or valleys within the target frame length.
  • the device further includes:
  • the prompt information sending module is configured to send prompt information to a preset device; wherein a communication connection is established between the preset device and the headset.
  • an earphone in another embodiment, includes a housing and a controller, a feedback microphone, a feedforward microphone and a speaker provided on the housing;
  • the feedforward microphone is connected to the controller and used to collect audio data outside the ear canal and send it to the controller;
  • the feedback microphone is connected to the controller and used to collect audio data in the ear canal and send it to the controller;
  • the controller includes a memory and a processor.
  • the memory stores executable computer instructions.
  • the processor can call the computer instructions stored in the memory to execute the method described in any of the above embodiments.
  • a computer storage medium stores an executable program; after the executable program is executed by a processor, the method described in any of the above embodiments can be implemented.
  • FIG. 7 is a block diagram of an electronic device 800 according to an exemplary embodiment.
  • the electronic device 800 may include one or more of the following components: a processing component 802 , a memory 804 , a power supply component 806 , a multimedia component 808 , an audio component 810 , an input/output (I/O) interface 812 , and a sensor component 814 , and communication component 816.
  • Processing component 802 generally controls the overall operations of electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations.
  • the processing component 802 may include one or more processors 820 to execute instructions to generate all or part of the steps of the methods described above.
  • processing component 802 may include one or more modules that facilitate interaction between processing component 802 and other components.
  • processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.
  • Memory 804 is configured to store various types of data to support operations at electronic device 800 . Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phonebook data, messages, pictures, videos, etc.
  • Memory 804 may 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 (EEPROM), Programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read-only memory
  • EEPROM erasable programmable read-only memory
  • EPROM Programmable read-only memory
  • PROM programmable read-only memory
  • ROM read-only memory
  • magnetic memory flash memory, magnetic or optical disk.
  • Power supply component 806 provides power to various components of electronic device 800 .
  • Power supply components 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 800 .
  • Multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user.
  • 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 touch screen 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 sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide action.
  • multimedia component 808 includes a front-facing camera and/or a rear-facing camera.
  • the front camera and/or the rear camera may receive external multimedia data.
  • Each front-facing camera and rear-facing camera can be a fixed optical lens system or have a focal length and optical zoom capabilities.
  • Audio component 810 is configured to output and/or input audio signals.
  • audio component 810 includes a microphone (MIC) configured to receive external audio signals when electronic device 800 is in operating modes, such as call mode, recording mode, and voice recognition mode. The received audio signal may be further stored in memory 804 or sent via communication component 816 .
  • audio component 810 also includes a speaker for outputting audio signals.
  • the I/O interface 812 provides an interface between the processing component 802 and a peripheral interface module, which may be a keyboard, a click wheel, a button, etc. These buttons may include, but are not limited to: Home button, Volume buttons, Start button, and Lock button.
  • Sensor component 814 includes one or more sensors for providing various aspects of status assessment for electronic device 800 .
  • the sensor component 814 can detect the open/closed state of the device 800, the relative positioning of components, such as the display and keypad of the electronic device 800.
  • the sensor component 814 can also detect the electronic device 800 or a component of the electronic device 800. changes in position, the presence or absence of user contact with the electronic device 800 , the orientation or acceleration/deceleration of the electronic device 800 and changes in the temperature of the electronic device 800 .
  • Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact.
  • Sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications.
  • the sensor component 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
  • Communication component 816 is configured to facilitate wired or wireless communication between electronic device 800 and other devices.
  • the electronic device 800 can access a wireless network based on a communication standard, such as WiFi, 4G or 5G, or a combination thereof.
  • the communication component 816 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • the communications component 816 also includes a near field communications (NFC) module to facilitate short-range communications.
  • NFC near field communications
  • the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency identification
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • electronic device 800 may be configured by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A programmable gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation is used to perform the above method.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs digital signal processing devices
  • PLDs programmable logic devices
  • FPGA field programmable A programmable gate array
  • controller microcontroller, microprocessor or other electronic component implementation is used to perform the above method.
  • a non-transitory computer-readable storage medium including instructions such as a memory 804 including instructions, executable by the processor 820 of the electronic device 800 to generate the above method is also provided.
  • the non-transitory computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Physics & Mathematics (AREA)
  • Dentistry (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Physiology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Headphones And Earphones (AREA)

Abstract

本申请公开一种跌倒检测方法、装置、耳机及存储介质。跌倒检测方法应用于耳机，耳机包括反馈麦克风，方法包括:通过反馈麦克风采集耳道内的音频信号，得到耳道音频信号，耳道音频信号包括:当耳机处于被用户佩戴的状态下，用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号（S100）；对耳道音频信号进行特征提取，得到音频信号特征参数（S200）；根据预设频率范围对耳道音频信号进行滤波，得到预设频率范围的周期信号，周期信号中包括波形的峰值或谷值（S300）；根据音频信号特征参数以及峰值或者谷值的数量，生成待识别特征（S400）；将待识别特征输入至跌倒检测模型，得到检测结果，检测结果至少用于表示用户发生跌倒（S500）。

Description

跌倒检测方法、装置、耳机及存储介质 技术领域
本公开涉及信息处理技术领域但不限于信息处理技术领域，尤其涉及一种跌倒检测方法、装置、耳机及存储介质。
背景技术
随着技术的发展，在各个应用场景中出现了越来越多的电子设备，不同的电子设备在对应的应用场景中可以实现不同的功能。
随着健康监测设备的广泛应用，通过健康监测设备可以检测到被检测对象的健康状态。例如，对目标人群的跌倒检测，通过在目标人群身体上佩戴相应的传感器，或者让目标人群穿戴具有相应传感器的服饰等，根据传感器检测的信号确定目标人群是否跌倒。目标人群可以包括老人和行动不便的人群等。
发明内容
本公开实施例提供一种跌倒检测方法、装置、耳机及存储介质。
本公开实施例第一方面提供一种跌倒检测方法，应用于耳机，所述耳机包括反馈麦克风，所述方法包括：
通过所述反馈麦克风采集耳道内的音频信号，得到耳道音频信号；其中，所述耳道音频信号包括：当所述耳机处于被用户佩戴的状态下，所述用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号；
对所述耳道音频信号进行特征提取，得到音频信号特征参数；
根据预设频率范围对所述耳道音频信号进行滤波，得到所述预设频率范围的周期信号；其中，所述周期信号中包括波形的峰值或谷值；
根据所述音频信号特征参数以及所述峰值或者所述谷值的数量，生成待识别特征；
将所述待识别特征输入至跌倒检测模型，得到检测结果；其中，所述检测结果用于表示所述用户发生跌倒。
本公开实施例第二方面提供一种跌倒检测装置，应用于耳机，所述耳机包括反馈麦克风，所述装置包括：
耳道音频信号检测模块，被配置为通过所述反馈麦克风采集耳道内的音频信号，得到耳道音频 信号；其中，所述耳道音频信号包括：当所述耳机处于被用户佩戴的状态下，所述用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号；
音频信号特征参数获取模块，被配置为对所述耳道音频信号进行特征提取，得到音频信号特征参数；
周期信息确定模块，被配置为根据预设频率范围对所述耳道音频信号进行滤波，得到所述预设频率范围的周期信号；其中，所述周期信号中包括波形的峰值或谷值；
待识别特征生成模块，被配置为根据所述音频信号特征参数以及所述峰值或所述谷值的数量，生成待识别特征；
检测模块，被配置为将所述待识别特征输入至跌倒检测模型，得到检测结果；其中，所述检测结果用于表示所述用户发生跌倒。
本公开实施例第三方面提供一种耳机，所述耳机包括壳体以及设置于所述壳体上的控制器、反馈麦克风、前馈麦克风和扬声器；所述前馈麦克风与所述控制器连接，用于采集耳道外音频数据并发送给所述控制器；所述反馈麦克风与所述控制器连接，用于采集耳道内音频数据并发送给所述控制器；所述控制器包括存储器和处理器，所述存储器上存储有可执行的计算机指令，所述处理器能够调用所述存储器上存储的计算机指令，以执行上述任意一实施例所述的方法。
本公开实施例第四方面提供一种计算机存储介质，所述计算机存储介质存储有可执行程序；所述可执行程序被处理器执行后，能够实现前述的第一方面提供的跌倒检测方法。
本公开实施例提供的跌倒检测方法可以应用于耳机，通过耳机即可确定出用户的是否跌倒，无需其他检测传感器，从而提高了检测用户跌倒的便利性，提高了用户的使用体验。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本公开实施例。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本发明实施例，并与说明书一起用于解释本发明实施例的原理。
图1是根据一示例性实施例示出的一种耳机的示意图；
图2是根据一示例性实施例示出的一种耳机的示意图；
图3是根据一示例性实施例示出的一种耳机处于被用户佩戴状态时的示意图；
图4是根据一示例性实施例示出的另一种跌倒检测方法的示意图；
图5是根据一示例性实施例示出的一种周期信号的示意图；
图6是根据一示例性实施例示出的一种跌倒检测装置的示意图；
图7是根据一示例性实施例示出的一种电子设备的结构示意图。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本发明实施例相一致的所有实施方式。相反，它们仅是本发明实施例的一些方面相一致的装置和方法的例子。
在本公开实施例使用的术语是仅仅出于描述特定实施例的目的，而非旨在限制本公开实施例。在本公开所使用的单数形式的“一种”、“所述”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义。还应当理解，本文中使用的术语“和/或”是指并包含一个或多个相关联的列出项目的任何或所有可能组合。
应当理解，尽管在本公开实施例可能采用术语第一、第二、第三等来描述各种信息，但这些信息不应限于这些术语。这些术语仅用来将同一类型的信息彼此区分开。例如，在不脱离本公开实施例范围的情况下，第一信息也可以被称为第二信息，类似地，第二信息也可以被称为第一信息。取决于语境，如在此所使用的词语“如果”可以被解释成为“在……时”或“当……时”或“响应于确定”。
通常情况下，对老人或者腿部功能出现异常等行动不便的人的人群进行跌倒检测，是通过多个传感器进行检测的。例如，在这些人群的身体上部署多个加速度传感器，或者这些人群穿戴配备有加速度传感器的服饰等，根据加速度传感器检测的信号的变化信息确定相关用户是否跌倒。
由于这些用户本身行动不便，通过该方法在这些用户身上再增加多个传感器进而会对这些用户产生影响，带来不适感并且导致行动更加不便。
参考图1，其示出了本公开实施例提供的一种跌倒检测方法的示意图，该方法至少可以应用于耳机，该耳机至少可以包括反馈麦克风。
如图1所示，该方法包括：
步骤S100，通过反馈麦克风采集耳道内的音频信号，得到耳道音频信号；其中，耳道音频信号包括：当耳机处于被用户佩戴的状态下，用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号。
步骤S200，对耳道音频信号进行特征提取，得到音频信号特征参数。
步骤S300，根据预设频率范围对耳道音频信号进行滤波，得到预设频率范围的周期信号；其中，周期信号中包括波形的峰值或谷值。
步骤S400，根据音频信号特征参数以及峰值或者谷值的数量，生成待识别特征。
步骤S500，将待识别特征输入至跌倒检测模型，得到检测结果；其中，检测结果至少用于表示用户发生跌倒。
耳机可以是不同形状的耳机，包括入耳式、半入耳式和头戴式等不同形式的耳机。耳机的通信方式可以包括有线耳机和无线耳机，无线耳机可以包括蓝牙耳机，例如真无线立体声耳机(True Wireless Stereo，TWS)。耳机还可以包括助听器等具有反馈麦克风并且能够实现该方案的设备。
耳机中的反馈麦克风可以位于耳机的出音通道附近，耳机处于佩戴状态时，反馈麦克风位于耳道内，可以采集耳道内的音频信号，例如入耳式耳机。在耳机为其他形式的耳机时，如半入耳式耳机和头戴式耳机等，在耳机处于佩戴状态时反馈麦克风能够采集耳道内的音频信号即可。
参考图2，为一种耳机的示意图，包括反馈麦克风A，在耳机处于佩戴状态时，该反馈麦克风A可以位于耳道内。该耳机还可以包括前馈麦克风B，前馈麦克风B可以位于耳机柄上，在耳机处于佩戴状态时，前馈麦克风位于耳道外，可以采集外部环境的环境音频信号。耳机还可以包括通话麦克风C，在通话状态下采集用户发出的音频信号。反馈麦克风A比前馈麦克风B采集到的音频信号的信噪比更高，所以通过反馈麦克风A采集到噪声更少、质量更高的耳道音频信号。
在耳机为头戴式耳机时，也会形成一定程度的堵耳效应，在耳机处于佩戴状态时，反馈麦克风A可以位于耳道外或者朝向耳道，能够采集耳道音频信号即可。
参考图3，为一种耳机处于被用户佩戴状态时的示意图，耳机1对耳道2形成一定程度的堵塞，形成堵耳效应。反馈麦克风采集的耳道音频信号包括：当耳机处于被用户佩戴的状态下，用户在呼吸过程中产生的震动通过骨传导方式被传递至耳道而产生的音频信号，即音频信号3。
对于步骤S100，在耳机处于佩戴状态时，对耳道产生一定的堵塞，形成一定程度的堵耳效应，产生原因是有部分声音会经人的骨头传导到内耳，例如，由于走路时脚部会与地面接触产生震动，震动通过骨传导的方式传递至耳道的音频信号。在耳机未处于佩戴状态时，骨传导来的声音一部分经外耳向外扩散，但是在耳机处于佩戴状态时，耳道在一定程度上被堵住，减少了骨传导来的声音通过耳道向外扩散的扩散量，形成一定程度的堵耳效应，也称闭塞效应。闭塞效应产生的声音特性表现为低频信号加强、高频信号衰弱。
由于耳机对耳道产生一定的堵塞，形成不同程度的堵耳效应，在产生堵耳效应后，耳机阻挡了外界音频信号进入耳道，减少了外界音频信号对耳道内音频信号的影响。反馈麦克风可以采集耳道内的音频信号得到呼吸音频信号，该呼吸音频信号包括当耳机处于被用户佩戴的状态下，用户在走路时脚部会与地面接触产生的震动通过骨传导方式被传递至耳道而产生的音频信号。
在形成一定程度的堵耳效应后，用户在走路时脚部会与地面接触产生的震动可以通过骨传导至耳道，产生音频信号，堵耳效应可以放大该音频信号，从而可以便于反馈麦克风采集耳道内的音频信号，得到耳道音频信号。
对于步骤S200，在得到耳道音频信号后，可以对耳道音频信号进行特征提取，得到音频信号特征参数。特征提取方式可以包括多种，例如通过特征提取算法提取相应的特征，得到的音频信号特征参数可以是梅尔谱系数和梅尔倒谱系数(MFCC)等，梅尔谱系数和梅尔倒谱系数都可以是40维的特征参数。当然还可以是耳道音频信号的其他特征。
在耳道音频信号不同时，对应的音频信号特征也会不同，不同的耳道音频信号具有各自的音频信号特征。在用户跌倒时身体与地面接触产生的震动通过骨传导方式被传递至耳道而产生的音频信号的音频信号特征，不同于用户在非跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳 道而产生的音频信号的音频信号特征，也不同于其他震动通过骨传导方式被传递至耳道而产生的音频信号的音频信号特征，例如，说话声、外界环境声音的信号的音频信号特征。
对于步骤S300，在得到耳道音频信号后，可以根据预设频率范围对耳道音频信号进行滤波，得到预设频率范围的周期信号，在该周期信号中包括波形的峰值。预设频率范围可以根据实际的使用需求确定，也可以是预设的。也可以根据预设数量的用户走路的步频确定。例如，预设频率范围可以是1Hz至50Hz，根据预设频率范围对耳道音频信号进行低通滤波。
根据预设频率范围，将耳道音频信号中预设频率范围之外的音频信号滤除，得到频率在预设频率范围之内的音频信号，预设频率范围之内的音频信号可以是周期信号，该周期信号可以是时域内的信号。
在该周期信号以波形的形式表示，包括时间和振幅的关系，根据周期信号可以确定出波形中波峰的峰值。波形中波峰的峰值表示在该峰值对应的时间用户的脚部与地面接触，一个峰值表示脚部与地面接触一次。根据峰值的数量可以确定出用户所走的步数。
根据周期信号可以确定出波形中波谷的谷值。波形中波谷的谷值表示在该谷值对应的时间用户的脚部与地面接触，一个谷值表示脚部与地面接触一次。根据谷值的数量可以确定出用户所走的步数。
步骤S200和步骤S300之间并没有必然的先后顺序关系，可以先执行步骤S200，也可以先执行步骤S300。
对于步骤S400，在确定出峰值或者谷值的数量以及音频信号特征参数后，可以根据峰值的数量和音频信号特征参数生成待识别特征，也可以根据谷值的数量和音频信号特征参数生成待识别特征。例如可以将峰值的数量和音频信号特征参数共同作为一个待识别特征，待识别特征中包括峰值的数量和音频信号特征参数两个维度的信息，用于确定检测结果。还可以将谷值的数量和音频信号特征参数共同作为一个待识别特征，待识别特征中包括谷值的数量和音频信号特征参数两个维度的信息，用于确定检测结果。
峰值的数量和/或音频信号特征参数不同时，得到的待识别特征不同。谷值的数量和/或音频信号特征参数不同时，得到的待识别特征不同。在峰值的数量和音频信号特征参数中的至少一个发生变化时，待识别特征则发生变化。谷值的数量和音频信号特征参数中的至少一个发生变化时，待识别特征则发生变化。这样可以减少在峰值的数量和音频信号特征参数同时发生变化时，或谷值的数量和音频信号特征参数同时发生变化时，得到的待识别特征不变的情况，从而可以提高检测结果的准确度。
对于步骤S500，将待识别特征输入至跌倒检测模型，得到检测结果。该检测结果至少用于表示用户发生跌倒。跌倒检测模型为提前训练完成的检测模型。
由于当耳机处于被用户佩戴的状态下，用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号的音频信号特征，不同于其他音频信号的音频信号特征，并且对应的周期信号中峰值或者谷值的数量也不同，所以将待识别特征输入至跌倒检测模型可以确定出用 户是否跌倒。
当然还可以通过其他方式确定，如映射表等，映射表中包括待识别特征和检测结果之间的映射关系，通过查找映射表，根据待识别特征可以确定检测结果。
本公开示例通过耳机可以获取用户的耳道音频信号，然后对耳道音频信号进行处理，并利用跌倒检测模型检测用户是否跌倒。利用已有的内置麦克风，无需新的硬件成本，无需使用其他各种传感器等监测设备，通过耳机即可确定出用户是否跌倒。降低了检测用户跌倒的难度和不便性，提高了检测用户跌倒的便利性，同时减少了在检测用户跌倒的过程给用户带来的不适感和不便性，提高了用户的使用体验。
在另一实施例中，在耳道音频信号中不包括当所述耳机处于被用户佩戴的状态下，所述用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号时，则通过步骤S200至S500的步骤，可以得到表示用户未跌倒的检测结果，此时可以输出该检测结果。
在一个实施例中，参考图4，图4为另一种跌倒检测方法的示意图。该方法还包括：
步骤S10，根据预设步数对应的时长确定至少一帧周期信号的目标帧长。
步骤S20，在目标帧长内确定出各帧周期信号中包括的峰值或者谷值的数量；其中，每一步对应一个峰值或谷值。
在生成待识别特征之前，还需求确定出峰值或者谷值的数量，可以根据预设步数对应的时长确定出至少一帧周期信号的目标帧长。预设步数可以根据实际的应用需求确定，例如可以是两步的时长，三步的时长或四步的时长等。预设步数对应的时长可以是以秒为单位，将预设步数对应的时长确定为一帧周期信号的目标帧长，也可以将预设步数对应的时长确定为N帧周期信号的目标帧长。在该实施例中，以N等于1，即预设步数对应的时长确定一帧周期信号的目标帧长为例，这样可以提高确定峰值或谷值的数量的准确度和便利性。
在确定目标帧长后，可以在目标帧长内确定出各帧周期信号中包括的峰值或者谷值的数量。由于周期信号以波形的形式表示，所以也可以根据周期信号确定出波形的波峰或波谷，根据波峰即可确定出波峰的峰值，根据波谷接口确定出波谷的谷值。根据目标帧长，确定出目标帧长内峰值或谷值的数量。每一步对应一个峰值或谷值，由于用户的脚部与地面接触时产生震动，音频信号的强度比脚部未与地面接触时对应的音频信号的强度大，所以目标帧长内峰值或谷值的数量可以表示用户所走的步数，即用户的脚部与地面的接触次数。
通过该方法可以确定出每一帧周期信号内对应的峰值或谷值的数量。
确定峰值的数量可以通过峰值检测算法对各帧周期信号进行检测，确定出各帧周期信号中的峰值。谷值的数量可以通过谷值检测算法对各帧周期信号进行检测，确定出各帧周期信号中的谷值。峰值检测算法可以检测出周期信号中波形的波峰，进而确定出对应的峰值，谷值检测算法可以检测出周期信号中波形的波谷，进而确定出对应的谷值。
在另一实施例中，步骤S400，可以根据一帧周期信号内峰值的数量生成待识别特征，也可以根据多帧周期信号内峰值的数量生成待识别特征。例如，可以根据多帧连续的周期信号内峰值的数量 的平均值和音频信号特征参数生成待识别特征。
根据多帧周期信号内峰值的数量生成待识别特征，这样可以减少个别少数帧的周期信号内峰值出现异常时对峰值的数量产生的影响，从而减少对待识别特征的影响，减少对检测结果的影响，提高检测结果的准确度。
参考图5，图5为一种周期信号的示意图。图5示出的周期信号为根据预设频率范围，对在未跌倒的情况下，正常走路对应音频信号进行滤波后得到的周期信号。图5中每个目标帧长都为两步对应的时长，时长为1秒。在振幅的正方向上，每个周期信号在目标帧长内包括三个峰值，以最左边一个目标帧长对应的周期信号为例，每个峰值表示脚部与地面接触，表示用户在通过脚部进行移动，最左边一个峰值至中间一个峰值的时长则为第一步的时长，中间峰值至最右边一个峰值的时长为下一步的时长。
图5中每个目标帧长内对应的波峰的峰值或者波谷的谷值用圆点表示，通过相应的检测算法可以确定出圆点表示的峰值或者谷值。
在另一实施例中，对于耳道音频信号中包括当所述耳机处于被用户佩戴的状态下，所述用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号时，由于跌倒时身体的多个部位会与地面接触碰撞，则在目标帧长每个周期信号中内包括峰值或谷值的数量多于未跌倒正常走路时对应峰值或谷值的数量。
根据图5可以得出，在未跌倒正常走路时对应的波峰或者波谷呈周期性，具有一定的规律，各个峰值之间的差值保持在一定的范围之内，各个峰值的大小相近，不会出现忽大忽小的情况。在跌倒时，身体多个不同部位与地面碰撞对应的波峰或者波谷则呈不规则状态，对应的峰值或者谷值也是不规则的。不同部位与地面的碰撞力度也不同，导致对应波峰的峰值或者波谷的谷值也不同。
在另一实施例中，跌倒检测模型为预先采用机器学习的方式基于跌倒信息训练样本集对初始神经网络模型进行训练得到。初始神经网络模型的结构并不进行限定，在通过机器学习的方式通过训练样本集进行训练后，能够根据待识别特征输出识别结果即可。
在一个实施例中，训练样本集包括正样本集，正样本集包括多个正样本。每个正样本包括：目标碰撞的音频特征、第一数量和第一标签。
目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在跌倒时身体与地面碰撞产生碰撞时，通过反馈麦克风采集耳道内的音频信号得到的音频特征。在耳机中反馈麦克风采集到耳道内低音频信号后，耳机可以对采集的音频信号进行特征提取，得到目标碰撞的音频特征。
第一数量为：根据预设频率范围对目标碰撞的音频特征进行滤波后，得到的预设频率范围的周期信号中包括的波形中峰值或谷值的数量。
第一标签用于表示目标碰撞的音频特征和第一数量对应所述初始神经网络模型的输出。
将正样本集中的各个正样本输入至初始神经网络模型中，第一标签作为初始神经网络的输出，训练初始神经网络模型，得到跌倒检测模型。正样本集中正样本的数量可以根据实际需求进行确定，数量越多，训练得到的跌倒检测模型的检测准确度越高。
目标碰撞的音频特征可以包括梅尔谱特征参数和梅尔倒谱特征参数等，梅尔谱系数和梅尔倒谱系数都可以是40维的系数。
在另一实施例中，不同正样本包括的目标碰撞的音频特征，对应的用户跌倒时身体与地面发生碰撞的姿势和/或碰撞次数不同。
每个正样本可以是不同用户在不同的跌倒姿势时，身体的不同部位与地面发生碰撞的音频信号的音频特征，并且不同的正样本包括的目标碰撞的音频特征对应的用户跌倒时身体与地面发生碰撞的姿势、与地面发生碰撞的部位和/或碰撞次数不同。
每个正样本也可以是同一用户在不同的跌倒姿势时，身体的不同部位与地面发生碰撞的音频信号的音频特征，并且不同的正样本包括的目标碰撞的音频特征对应的用户跌倒时身体与地面发生碰撞的姿势、与地面发生碰撞的部位和/或碰撞次数不同。
例如，正样本1包括的音频特征为用户1以姿态1跌倒时身体与地面发生碰撞的音频特征，用户1以姿态1跌倒时身体不同部位与地面发生碰撞的次数为次数1，与地面发生碰撞的部位边框手部和膝盖部位。正样本2包括的音频特征为用户2以姿态2跌倒时身体与地面发生碰撞的音频特征，用户2以姿态2跌倒时身体不同部位与地面发生碰撞的次数为次数2，与地面发生碰撞的部位包括手部和臀部。正样本3包括的音频特征为用户1以姿态2跌倒时身体与地面发生碰撞的音频特征，用户1以姿态2跌倒时身体不同部位与地面发生碰撞的次数为次数3，与地面发生碰撞的部位包括背部和头部。
在另一实施例中，第一数量为在目标帧长内的峰值或谷值的数量。正样本中对应第一数量和生成待识别特征的峰值或谷值的数量都是在相同的帧长内确定的，这样可以减少变量，提高检测的准确度。
在一个实施例中，训练样本集还包括负样本集，负样本集包括多个负样本。每个负样本包括：非目标碰撞的音频特征、第二数量和第二标签。
非目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在除了跌倒时身体与地面碰撞产生碰撞以外的情况下，通过反馈麦克风采集耳道内的音频信号得到的音频特征。
第二数量为：根据预设频率范围对非目标碰撞的音频特征进行滤波后，得到的预设频率范围的周期信号中包括的波形的峰值或谷值的数量。
第二标签用于表示非目标碰撞的音频特征和第二数量对应的初始神经网络模型的输出。
负样本包括的音频特征与正样本包括的音频特征不同，正样本包括的是反馈麦克风采集的各种跌倒状态下身体与地面碰撞的音频信号的音频特征，负样本包括的是反馈麦克风采集的非跌倒状态下，耳道内的各种音频信号的音频特征。非跌倒状态下，耳道内的音频信号可以包括环境音频、说话的音频以及用户的其他交互操作产生的音频的音频特征，这些音频特征与正样本包括的目标碰撞的音频特征不同。
将负样本集中的各个负样本输入至初始神经网络模型中，第二标签作为初始神经网络的输出，训练初始神经网络模型，得到跌倒检测模型。负样本集中负样本的数量可以根据实际需求进行确定， 数量越多，训练得到的跌倒检测模型的检测准确度越高。
非目标碰撞的音频特征可以包括梅尔谱特征参数和梅尔倒谱特征参数等，梅尔谱系数和梅尔倒谱系数都可以是40维的系数。
在另一实施例中，第二数量为在目标帧长内的峰值或谷值的数量。负样本中对应第二数量和生成待识别特征的峰值或谷值的数量都是在相同的帧长内确定的，这样可以减少变量，提高检测的准确度。
利用正样本和负样本训练初始神经网络模型，得到跌倒检测模型，提高了跌倒检测模型的检测能力，得到的检测结果更加准确。
在另一实施例中，跌倒检测方法还包括：
在检测结果为表示用户跌倒的检测结果时，向与耳机建立有通信连接的预设设备发送提示信息。预设设备可以是手机、平板电脑等设备，还可以是与被检测用户具有社交关系的用户持有的设备。例如，用户为老人，则预设设备可以是看护人员的电子设备
向与耳机建立有通信连接的预设设备发送提示信息，以便通知相关人员，从而有利对用户进行帮助。
提示信息可以弹窗消息、声音提示信息或者短消息等。
在另一实施例中，参考图6，为一种跌倒检测装置的示意图，其中，应用于耳机，耳机包括反馈麦克风，所述装置包括：
耳道音频信号检测模块1，被配置为通过所述反馈麦克风采集耳道内的音频信号，得到耳道音频信号；其中，所述耳道音频信号包括：当所述耳机处于被用户佩戴的状态下，所述用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号；
音频信号特征参数获取模块2，被配置为对所述耳道音频信号进行特征提取，得到音频信号特征参数；
周期信号确定模块3，被配置为根据预设频率范围对所述耳道音频信号进行滤波，得到所述预设频率范围的周期信号；其中，所述周期信号中包括波形的峰值或谷值；
待识别特征生成模块4，被配置为根据所述音频信号特征参数以及所述峰值或所述谷值的数量，生成待识别特征；
检测模块5，被配置为将所述待识别特征输入至跌倒检测模型，得到检测结果；其中，所述检测结果至少用于表示所述用户发生跌倒。
在另一实施例中，所述装置还包括：
目标帧长确定模块，被配置为根据预设步数对应的时长确定至少一帧所述周期信号的目标帧长；
数量确定模块，被配置为在所述目标帧长内确定出各帧所述周期信号中包括的所述峰值或者所述谷值的数量；其中，每一步对应一个所述峰值或所述谷值。
在另一实施例中，所述跌倒检测模型为预先采用机器学习的方式基于跌倒信息训练样本集对初始神经网络模型进行训练得到。
在另一实施例中，所述训练样本集包括正样本集，所述正样本集包括多个正样本；
每个所述正样本包括：目标碰撞的音频特征、第一数量和第一标签；
其中，所述目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在跌倒时身体与地面碰撞产生碰撞时，通过所述反馈麦克风采集耳道内的音频信号得到的音频特征；
所述第一数量为：根据所述预设频率范围对所述目标碰撞的音频特征进行滤波后，得到的所述预设频率范围的周期信号中包括的波形中峰值或谷值的数量；
所述第一标签用于表示所述目标碰撞的音频特征和所述第一数量对应所述初始神经网络模型的输出。
在另一实施例中，所述第一数量为在目标帧长内的峰值或谷值的数量。
在另一实施例中，不同所述正样本包括的目标碰撞的音频特征，对应的用户跌倒时身体与地面发生碰撞的姿势、与地面发生碰撞的部位和/或碰撞次数不同。
在另一实施例中，所述训练样本集还包括负样本集，所述负样本集包括多个负样本；
每个所述负样本包括：非目标碰撞的音频特征、第二数量和第二标签；
所述非目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在除了跌倒时身体与地面碰撞产生碰撞以外的情况下，通过所述反馈麦克风采集耳道内的音频信号得到的音频特征；
所述第二数量为：根据所述预设频率范围对所述非目标碰撞的音频特征进行滤波后，得到的所述预设频率范围的周期信号中包括的波形中峰值或谷值的数量；
所述第二标签用于表示所述非目标碰撞的音频特征和所述第二数量对应所述初始神经网络模型的输出。
在另一实施例中，所述第二数量为在目标帧长内的峰值或谷值的数量。
在另一实施例中，所述装置还包括：
提示信息发送模块，被配置为向预设设备发送提示信息；其中，所述预设设备与所述耳机之间建立有通信连接。
在另一实施例中，提供一种耳机，所述耳机包括壳体以及设置于所述壳体上的控制器、反馈麦克风、前馈麦克风和扬声器；
所述前馈麦克风与所述控制器连接，用于采集耳道外音频数据并发送给所述控制器；
所述反馈麦克风与所述控制器连接，用于采集耳道内音频数据并发送给所述控制器；
所述控制器包括存储器和处理器，所述存储器上存储有可执行的计算机指令，所述处理器能够调用所述存储器上存储的计算机指令，以执行上述任意一实施例所述的方法。
在另一实施例中，提供一种计算机存储介质，所述计算机存储介质存储有可执行程序；所述可执行程序被处理器执行后，能够实现上述任意一实施例所述的方法。
图7是根据一示例性实施例示出的一种电子设备800的框图。
参照图7，电子设备800可以包括以下一个或多个组件：处理组件802，存储器804，电源组件806，多媒体组件808，音频组件810，输入/输出(I/O)的接口812，传感器组件814，以及通信组 件816。
处理组件802通常控制电子设备800的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理组件802可以包括一个或多个处理器820来执行指令，以生成上述的方法的全部或部分步骤。此外，处理组件802可以包括一个或多个模块，便于处理组件802和其他组件之间的交互。例如，处理组件802可以包括多媒体模块，以方便多媒体组件808和处理组件802之间的交互。
存储器804被配置为存储各种类型的数据以支持在电子设备800的操作。这些数据的示例包括用于在电子设备800上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器804可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电源组件806为电子设备800的各种组件提供电力。电源组件806可以包括电源管理系统，一个或多个电源，及其他与为电子设备800生成、管理和分配电力相关联的组件。
多媒体组件808包括在所述电子设备800和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。所述触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与所述触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件808包括一个前置摄像头和/或后置摄像头。当电子设备800处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光学变焦能力。
音频组件810被配置为输出和/或输入音频信号。例如，音频组件810包括一个麦克风(MIC)，当电子设备800处于操作模式，如呼叫模式、记录模式和语音识别模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器804或经由通信组件816发送。在一些实施例中，音频组件810还包括一个扬声器，用于输出音频信号。
I/O接口812为处理组件802和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件814包括一个或多个传感器，用于为电子设备800提供各个方面的状态评估。例如，传感器组件814可以检测到设备800的打开/关闭状态，组件的相对定位，例如所述组件为电子设备800的显示器和小键盘，传感器组件814还可以检测电子设备800或电子设备800一个组件的位置改变，用户与电子设备800接触的存在或不存在，电子设备800方位或加速/减速和电子设备800的温度变化。传感器组件814可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件814还可以包括光传感器，如CMOS或CCD图像传感器，用于在成像应用 中使用。在一些实施例中，该传感器组件814还可以包括加速度传感器，陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件816被配置为便于电子设备800和其他设备之间有线或无线方式的通信。电子设备800可以接入基于通信标准的无线网络，如WiFi，4G或5G，或它们的组合。在一个示例性实施例中，通信组件816经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信组件816还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，电子设备800可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述方法。
在示例性实施例中，还提供了一种包括指令的非临时性计算机可读存储介质，例如包括指令的存储器804，上述指令可由电子设备800的处理器820执行以生成上述方法。例如，所述非临时性计算机可读存储介质可以是ROM、随机存取存储器(RAM)、CD-ROM、磁带、软盘和光数据存储设备等。
本领域技术人员在考虑说明书及实践这里公开的发明后，将容易想到本发明的其它实施方案。本公开旨在涵盖本发明的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本发明的一般性原理并包括本公开未公开的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本发明的真正范围和精神由下面的权利要求指出。
应当理解的是，本发明并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本发明的范围仅由所附的权利要求来限制。

Claims (20)

  1. 一种跌倒检测方法，应用于耳机，所述耳机包括反馈麦克风，所述方法包括：
    通过所述反馈麦克风采集耳道内的音频信号，得到耳道音频信号；其中，所述耳道音频信号包括：当所述耳机处于被用户佩戴的状态下，所述用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号；
    对所述耳道音频信号进行特征提取，得到音频信号特征参数；
    根据预设频率范围对所述耳道音频信号进行滤波，得到所述预设频率范围的周期信号；其中，所述周期信号中包括波形的峰值或谷值；
    根据所述音频信号特征参数以及所述峰值或者所述谷值的数量，生成待识别特征；
    将所述待识别特征输入至跌倒检测模型，得到检测结果；其中，所述检测结果至少用于表示所述用户发生跌倒。
  2. 根据权利要求1所述的方法，其中，所述方法还包括：
    根据预设步数对应的时长确定至少一帧所述周期信号的目标帧长；
    在所述目标帧长内确定出各帧所述周期信号中包括的所述峰值或者所述谷值的数量；其中，每一步对应一个所述峰值或所述谷值。
  3. 根据权利要求1或2所述的方法，其中，所述跌倒检测模型为预先采用机器学习的方式基于跌倒信息训练样本集对初始神经网络模型进行训练得到。
  4. 根据权利要求3所述的方法，其中，所述训练样本集包括正样本集，所述正样本集包括多个正样本；
    每个所述正样本包括：目标碰撞的音频特征、第一数量和第一标签；
    其中，所述目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在跌倒时身体与地面碰撞产生碰撞时，通过所述反馈麦克风采集耳道内的音频信号得到的音频特征；
    所述第一数量为：根据所述预设频率范围对所述目标碰撞的音频特征进行滤波后，得到的所述预设频率范围的周期信号中包括的波形中峰值或谷值的数量；
    所述第一标签用于表示所述目标碰撞的音频特征和所述第一数量对应所述初始神经网络模型的输出。
  5. 根据权利要求4所述的方法，其中，所述第一数量为在目标帧长内的峰值或谷值的数量。
  6. 根据权利要求4所述的方法，其中，不同所述正样本包括的目标碰撞的音频特征，对应的用户跌倒时身体与地面发生碰撞的姿势、与地面发生碰撞的部位和/或碰撞次数不同。
  7. 根据权利要求3所述的方法，其中，所述训练样本集还包括负样本集，所述负样本集包括多个负样本；
    每个所述负样本包括：非目标碰撞的音频特征、第二数量和第二标签；
    所述非目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在除了跌倒时身体与地面碰 撞产生碰撞以外的情况下，通过所述反馈麦克风采集耳道内的音频信号得到的音频特征；
    所述第二数量为：根据所述预设频率范围对所述非目标碰撞的音频特征进行滤波后，得到的所述预设频率范围的周期信号中包括的波形中峰值或谷值的数量；
    所述第二标签用于表示所述非目标碰撞的音频特征和所述第二数量对应所述初始神经网络模型的输出。
  8. 根据权利要求7所述的方法，其中，所述第二数量为在目标帧长内的峰值或谷值的数量。
  9. 根据权利要求1所述的方法，其中，所述方法还包括：
    向预设设备发送提示信息；其中，所述预设设备与所述耳机之间建立有通信连接。
  10. 一种跌倒检测装置，应用于耳机，所述耳机包括反馈麦克风，所述装置包括：
    耳道音频信号检测模块，被配置为通过所述反馈麦克风采集耳道内的音频信号，得到耳道音频信号；其中，所述耳道音频信号包括：当所述耳机处于被用户佩戴的状态下，所述用户在跌倒时身体与地面碰撞产生的震动通过骨传导方式被传递至耳道而产生的音频信号；
    音频信号特征参数获取模块，被配置为对所述耳道音频信号进行特征提取，得到音频信号特征参数；
    周期信号确定模块，被配置为根据预设频率范围对所述耳道音频信号进行滤波，得到所述预设频率范围的周期信号；其中，所述周期信号中包括波形的峰值或谷值；
    待识别特征生成模块，被配置为根据所述音频信号特征参数以及所述峰值或者所述谷值的数量，生成待识别特征；
    检测模块，被配置为将所述待识别特征输入至跌倒检测模型，得到检测结果；其中，所述检测结果至少用于表示所述用户发生跌倒。
  11. 根据权利要求10所述的装置，其中，所述装置还包括：
    目标帧长确定模块，被配置为根据预设步数对应的时长确定至少一帧所述周期信号的目标帧长；
    数量确定模块，被配置为在所述目标帧长内确定出各帧所述周期信号中包括的所述峰值或者所述谷值的数量；其中，每一步对应一个所述峰值或所述谷值。
  12. 根据权利要求10或11所述的装置，其中，所述跌倒检测模型为预先采用机器学习的方式基于跌倒信息训练样本集对初始神经网络模型进行训练得到。
  13. 根据权利要求12所述的装置，其中，所述训练样本集包括正样本集，所述正样本集包括多个正样本；
    每个所述正样本包括：目标碰撞的音频特征、第一数量和第一标签；
    其中，所述目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在跌倒时身体与地面碰撞产生碰撞时，通过所述反馈麦克风采集耳道内的音频信号得到的音频特征；
    所述第一数量为：根据所述预设频率范围对所述目标碰撞的音频特征进行滤波后，得到的所述预设频率范围的周期信号中包括的波形中峰值或谷值的数量；
    所述第一标签用于表示所述目标碰撞的音频特征和所述第一数量对应所述初始神经网络模型的 输出。
  14. 根据权利要求13所述的装置，其中，所述第一数量为在目标帧长内的峰值或谷值的数量。
  15. 根据权利要求13所述的装置，其中，不同所述正样本包括的目标碰撞的音频特征，对应的用户跌倒时身体与地面发生碰撞的姿势、与地面发生碰撞的部位和/或碰撞次数不同。
  16. 根据权利要求12所述的装置，其中，所述训练样本集还包括负样本集，所述负样本集包括多个负样本；
    每个所述负样本包括：非目标碰撞的音频特征、第二数量和第二标签；
    所述非目标碰撞的音频特征为耳机处于被用户佩戴的状态下，用户在除了跌倒时身体与地面碰撞产生碰撞以外的情况下，通过所述反馈麦克风采集耳道内的音频信号得到的音频特征；
    所述第二数量为：根据所述预设频率范围对所述非目标碰撞的音频特征进行滤波后，得到的所述预设频率范围的周期信号中包括的波形中峰值或谷值的数量；
    所述第二标签用于表示所述非目标碰撞的音频特征和所述第二数量对应所述初始神经网络模型的输出。
  17. 根据权利要求16所述的装置，其中，所述第二数量为在目标帧长内的峰值或谷值的数量。
  18. 根据权利要求10所述的装置，其中，所述装置还包括：
    提示信息发送模块，被配置为向预设设备发送提示信息；其中，所述预设设备与所述耳机之间建立有通信连接。
  19. 一种耳机，所述耳机包括壳体以及设置于所述壳体上的控制器、反馈麦克风、前馈麦克风和扬声器；
    所述前馈麦克风与所述控制器连接，用于采集耳道外音频数据并发送给所述控制器；
    所述反馈麦克风与所述控制器连接，用于采集耳道内音频数据并发送给所述控制器；
    所述控制器包括存储器和处理器，所述存储器上存储有可执行的计算机指令，所述处理器能够调用所述存储器上存储的计算机指令，以执行权利要求1至9中任意一项所述的方法。
  20. 一种计算机存储介质，所述计算机存储介质存储有可执行程序；所述可执行程序被处理器执行后，能够实现如权利要求1至9任一项提供的方法。
PCT/CN2022/099024 2022-06-15 2022-06-15 跌倒检测方法、装置、耳机及存储介质 WO2023240512A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202280004515.1A CN117597065A (zh) 2022-06-15 2022-06-15 跌倒检测方法、装置、耳机及存储介质
PCT/CN2022/099024 WO2023240512A1 (zh) 2022-06-15 2022-06-15 跌倒检测方法、装置、耳机及存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/099024 WO2023240512A1 (zh) 2022-06-15 2022-06-15 跌倒检测方法、装置、耳机及存储介质

Publications (1)

Publication Number Publication Date
WO2023240512A1 true WO2023240512A1 (zh) 2023-12-21

Family

ID=89192797

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/099024 WO2023240512A1 (zh) 2022-06-15 2022-06-15 跌倒检测方法、装置、耳机及存储介质

Country Status (2)

Country Link
CN (1) CN117597065A (zh)
WO (1) WO2023240512A1 (zh)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104065776A (zh) * 2014-04-03 2014-09-24 上海理工大学 一种弱势人群跌倒监测系统
EP3171612A1 (fr) * 2015-11-19 2017-05-24 Parrot Drones Casque audio à contrôle actif de bruit, contrôle anti-occlusion et annulation de l'atténuation passive, en fonction de la présence ou de l'absence d'une activité vocale de l'utilisateur de casque
CN110916675A (zh) * 2019-11-29 2020-03-27 歌尔科技有限公司 一种头戴式设备及其跌倒检测方法、装置
US20200236479A1 (en) * 2018-12-15 2020-07-23 Starkey Laboratories, Inc. Hearing assistance system with enhanced fall detection features
CN111447523A (zh) * 2020-03-31 2020-07-24 歌尔科技有限公司 耳机及其降噪方法、计算机可读存储介质
CN215647254U (zh) * 2021-05-27 2022-01-25 西安闻泰信息技术有限公司 Tws耳机及tws耳机系统
CN114267152A (zh) * 2021-12-17 2022-04-01 歌尔科技有限公司 防摔提醒方法、设备及计算机可读存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104065776A (zh) * 2014-04-03 2014-09-24 上海理工大学 一种弱势人群跌倒监测系统
EP3171612A1 (fr) * 2015-11-19 2017-05-24 Parrot Drones Casque audio à contrôle actif de bruit, contrôle anti-occlusion et annulation de l'atténuation passive, en fonction de la présence ou de l'absence d'une activité vocale de l'utilisateur de casque
US20200236479A1 (en) * 2018-12-15 2020-07-23 Starkey Laboratories, Inc. Hearing assistance system with enhanced fall detection features
CN110916675A (zh) * 2019-11-29 2020-03-27 歌尔科技有限公司 一种头戴式设备及其跌倒检测方法、装置
CN111447523A (zh) * 2020-03-31 2020-07-24 歌尔科技有限公司 耳机及其降噪方法、计算机可读存储介质
CN215647254U (zh) * 2021-05-27 2022-01-25 西安闻泰信息技术有限公司 Tws耳机及tws耳机系统
CN114267152A (zh) * 2021-12-17 2022-04-01 歌尔科技有限公司 防摔提醒方法、设备及计算机可读存储介质

Also Published As

Publication number Publication date
CN117597065A (zh) 2024-02-23

Similar Documents

Publication Publication Date Title
EP3081011B1 (en) Name-sensitive listening device
US11223915B2 (en) Detecting user's eye movement using sensors in hearing instruments
CN112087701B (zh) 用于风检测的麦克风的扬声器仿真
CN113905320B (zh) 为考虑语音检测而调节声音回放的方法和系统
CN110390953B (zh) 啸叫语音信号的检测方法、装置、终端及存储介质
CN116324969A (zh) 具有定位反馈的听力增强和可穿戴系统
CN112037825B (zh) 音频信号的处理方法及装置、存储介质
CN109256145B (zh) 基于终端的音频处理方法、装置、终端和可读存储介质
JP2017520179A (ja) 中心に位置するイヤピースを含むモバイルデバイス
CN113810828A (zh) 音频信号处理方法、装置、可读存储介质及耳机
WO2023240512A1 (zh) 跌倒检测方法、装置、耳机及存储介质
CN115550791A (zh) 音频处理方法、装置、耳机及存储介质
CN113596662B (zh) 啸叫声的抑制方法、啸叫声的抑制装置、耳机及存储介质
CN114040309B (zh) 风噪检测方法、装置、电子设备及存储介质
Bhowmik et al. Hear, now, and in the future: Transforming hearing aids into multipurpose devices
CN107677363A (zh) 噪音提示的方法及智能终端
WO2023240510A1 (zh) 呼吸监测方法、装置、耳机及存储介质
CN115278441A (zh) 语音检测方法、装置、耳机及存储介质
US10873816B2 (en) Providing feedback of an own voice loudness of a user of a hearing device
CN113450537B (zh) 跌倒检测方法、装置、电子设备和存储介质
WO2023245372A1 (zh) 计步方法、装置、耳机及存储介质
CN113825081B (zh) 一种基于掩蔽治疗系统的助听方法及装置
CN113660594B (zh) 一种助听系统自调节降噪方法及装置
US20230403523A1 (en) Method and system for fitting a hearing aid to a user
CN113766402B (zh) 一种提高环境适应性的助听方法及装置

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 202280004515.1

Country of ref document: CN

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22946206

Country of ref document: EP

Kind code of ref document: A1