WO2018103028A1 - 一种音频播放装置、系统及方法 - Google Patents

一种音频播放装置、系统及方法 Download PDF

Info

Publication number
WO2018103028A1
WO2018103028A1 PCT/CN2016/108971 CN2016108971W WO2018103028A1 WO 2018103028 A1 WO2018103028 A1 WO 2018103028A1 CN 2016108971 W CN2016108971 W CN 2016108971W WO 2018103028 A1 WO2018103028 A1 WO 2018103028A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
music
heart rate
motion state
acceleration data
Prior art date
Application number
PCT/CN2016/108971
Other languages
English (en)
French (fr)
Inventor
杨柯
陈法海
Original Assignee
深圳市汇顶科技股份有限公司
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 深圳市汇顶科技股份有限公司 filed Critical 深圳市汇顶科技股份有限公司
Priority to CN201680001779.6A priority Critical patent/CN106687958A/zh
Priority to PCT/CN2016/108971 priority patent/WO2018103028A1/zh
Publication of WO2018103028A1 publication Critical patent/WO2018103028A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
    • G06F16/63—Querying
    • G06F16/635—Filtering based on additional data, e.g. user or group profiles
    • G06F16/636—Filtering based on additional data, e.g. user or group profiles by using biological or physiological data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
    • G06F16/63—Querying
    • G06F16/638—Presentation of query results
    • G06F16/639—Presentation of query results using playlists

Definitions

  • Embodiments of the present invention relate to the field of audio playback technologies, and in particular, to an audio playback device, system, and method.
  • mainstream audio playback devices cannot automatically select or switch the type of music to be played for different motion states, and there is no music library that is subdivided according to the music rhythm matched by different motion states.
  • the user needs to manually select a music category or play. The specified song. If you can automatically recognize the user's motion state, and automatically play or switch the music that matches the user's current motion state, rhythm and heart rate, it will undoubtedly greatly improve the user's experience and meet the user's individual needs.
  • the embodiments of the present invention provide an audio playback device, system, and method, which are intended to solve at least one of the above technical problems in the prior art.
  • the technical solutions adopted by the embodiments of the present invention include:
  • An audio playback device configured to receive user heart rate and acceleration data transmitted by a headset, identify a user motion state according to the user heart rate and acceleration data, and transmit the user heart rate and the user motion state to a server, Receiving a music playlist transmitted by the server that matches the user's heart rate and the user's motion state, and playing the music in the music playlist.
  • the audio playback device includes:
  • a first communication module for communication interaction between the audio playback device and the earphone and the server;
  • a state recognition module configured to identify a corresponding user motion state according to user heart rate and acceleration data
  • Audio playback module for playing music corresponding to a music playlist transmitted by the server.
  • the state recognition module includes:
  • a data preprocessing unit configured to preprocess acceleration data transmitted by the earphone
  • Feature extraction unit configured to extract acceleration feature information in different motion states according to the pre-processed acceleration data
  • a first state identifying unit configured to train a state recognition model in different motion states according to the extracted acceleration feature information, and initially identify a current motion state of the user by using the state recognition model;
  • a second state recognition unit configured to calculate a user heart rate value according to the user's heart rate, identify a typical motion state of the user according to the relationship between the user heart rate value interval and the motion state, and identify the current motion state of the user according to the state recognition model.
  • the typical motion state of the user is compensated to obtain the final user motion state.
  • the technical solution adopted by the embodiment of the present invention further includes: preprocessing the acceleration data transmitted by the earphone by the data pre-processing unit, specifically: filtering the noise in the acceleration data transmitted by the earphone through a low-pass filter, and then using the third-order
  • the moving average filter further filters the random noise in the acceleration data filtered by the low-pass filter to obtain accurate acceleration data, and divides the accurate acceleration data through different time windows.
  • an audio playing system including an audio playing device, and an earphone and a server;
  • the earphone is configured to acquire user heart rate and acceleration data, and transmit the acquired user heart rate and acceleration data to the audio playback device;
  • the audio playback device is configured to identify a user motion state according to the user heart rate and acceleration data, and transmit the user heart rate and the user motion state to a server;
  • the server is configured to match a corresponding music database according to the user motion state, select a music playlist whose music rhythm matches the user's heart rate in the matched music database, and transmit the music playlist to the audio playback device Play it.
  • the technical solution adopted by the embodiment of the present invention further includes: the earphone comprises a heart rate sensor, and the heart rate sensor is configured to acquire a user heart rate.
  • the technical solution adopted by the embodiment of the present invention further includes: the earphone further includes an acceleration sensor, wherein the acceleration sensor is used to acquire acceleration data of the user; and the acceleration data acquired by the acceleration sensor includes three directions of x, y, and z. data.
  • the technical solution adopted by the embodiment of the present invention further includes: the earphone further includes a second communication module, wherein the second communication module is configured to transmit the user heart rate and acceleration data to an audio playback device connected to the earphone.
  • the server includes:
  • Music classification module for classifying music according to music type, language, sport state and music rhythm, and establishing a corresponding music database according to the classification result;
  • the music matching module is configured to receive a user heart rate and a user motion state transmitted by the audio playback device, match the corresponding music database according to the user motion state, and select a music playlist whose music rhythm matches the user's heart rate in the matched music database, and The music playlist is transmitted to an audio playback device.
  • the technical solution adopted by the embodiment of the present invention further includes: the music classification module classifies the music according to the music rhythm by: detecting the rhythm point and the rhythm value of the music from the PCM encoded music signal by the music rhythm detection algorithm, according to the music The rhythm points and rhythm values classify the music for each sport state.
  • Another technical solution adopted by the embodiment of the present invention is an audio playing method, which includes the following steps:
  • Step a receiving user heart rate and acceleration data transmitted by the headset;
  • Step b identifying a user motion state according to the user heart rate and acceleration data, and the user Heart rate and user motion status are transmitted to the server;
  • Step c receiving a music playlist transmitted by the server that matches the user's heart rate and the user's motion state, and playing the music in the music playlist.
  • the technical solution adopted by the embodiment of the present invention further includes: in the step a, a heart rate sensor and an acceleration sensor are respectively disposed in the earphone, wherein the heart rate sensor is used to acquire a user heart rate; and the acceleration sensor is used to acquire x, Acceleration data in three directions y and z.
  • the identifying the user motion state according to the user heart rate and acceleration data includes:
  • Step b1 pre-processing acceleration data transmitted by the earphone
  • Step b2 extract acceleration characteristic information in different motion states according to the pre-processed acceleration data
  • Step b3 training the state recognition model in different motion states according to the extracted acceleration feature information, and initially identifying the current motion state of the user through the state recognition model;
  • Step b4 calculating a user heart rate value according to the user's heart rate, identifying a typical motion state of the user according to the relationship between the user heart rate value interval and the motion state, and identifying the current motion state of the user based on the state recognition model to the user The typical motion state is compensated for the final user motion state.
  • the technical solution adopted by the embodiment of the present invention further includes: in the step b1, the data pre-processing specifically includes: filtering, by using a low-pass filter, noise in the acceleration data transmitted by the earphone, and then using third-order moving average filtering Further filtering out randomness in the acceleration data filtered by the low-pass filter Noise, accurate acceleration data is obtained; and the accurate acceleration data is segmented through different time windows.
  • the data pre-processing specifically includes: filtering, by using a low-pass filter, noise in the acceleration data transmitted by the earphone, and then using third-order moving average filtering Further filtering out randomness in the acceleration data filtered by the low-pass filter Noise, accurate acceleration data is obtained; and the accurate acceleration data is segmented through different time windows.
  • the technical solution adopted by the embodiment of the present invention further includes: the step c further comprises: classifying the music according to the music type, the language, the motion state, and the music rhythm by the server, and establishing a corresponding music database according to the classification result.
  • the technical solution adopted by the embodiment of the present invention further includes: the classification manner of classifying the music according to the music rhythm is: detecting a rhythm point and a rhythm value of the music from the PCM encoded music signal by the music rhythm detection algorithm, according to the rhythm point of the music And the rhythm value classifies the music for each sport state.
  • the beneficial effects of the embodiments of the present invention are: the audio playing device, the system and the method of the embodiment of the present invention acquire the user heart rate and acceleration data in real time through the heart rate sensor and the acceleration sensor in the earphone, and the user heart rate and The acceleration data is transmitted to the audio playback device connected to the earphone, and the audio playback device accurately identifies the corresponding user motion state according to the user's heart rate and acceleration data; and transmits the user heart rate and the user motion state to the server, and the server transmits the user's motion state.
  • a music playlist that matches and the music rhythm matches the user's heart rate is played.
  • the invention can increase the pleasure during exercise and achieve better exercise effects without manual operation by the user.
  • the invention has less hardware, lower cost, simple design, convenient development, and expands the use function of the ordinary earphone, which is beneficial to improving the user experience and satisfying the individual needs of the user.
  • FIG. 1 is a schematic structural diagram of an audio playback device according to an embodiment of the present invention.
  • FIG. 2 is a schematic structural diagram of an audio playing system according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of data acquired by an acceleration sensor
  • Figure 4 is the acceleration data acquired during the walking process
  • FIG. 5 is a flowchart of an audio playing method according to an embodiment of the present invention.
  • the application device is installed in the audio playback device 100 of the embodiment of the present invention.
  • the application software can accurately identify the user's motion state according to the user's heart rate and acceleration data transmitted by the headset, and transmit the user's heart rate and the user's motion state to the server, and receive the transmission of the server.
  • a music playlist that matches the user's heart rate and user motion state, and plays the music in the music playlist.
  • the audio playback device 100 includes a first communication module 110, a state recognition module 120, and an audio playback module 130;
  • the first communication module 110 is used for communication interaction between the audio playback device 100 and the earphone and the server; Specifically, the method includes: receiving user heart rate and acceleration data transmitted by the earphone, transmitting the user motion state recognized by the user heart rate and the state recognition module 120 to the server 300, and receiving a music playlist transmitted by the server and matching the user heart rate and the user motion state;
  • the communication manner of the first communication module 110 includes, but is not limited to, wired, Bluetooth, WiFi, and the like.
  • the state recognition module 120 is configured to identify a corresponding user motion state according to the user heart rate and acceleration data, and transmit the user heart rate and the user motion state to the server 300 by the first communication module 110.
  • the state recognition module 120 includes a data pre-processing unit 121, a feature extraction unit 122, a first state recognition unit 123, and a second state recognition unit 124;
  • the data pre-processing unit 121 is configured to pre-process the acceleration data acquired by the earphone; the pre-processing specifically includes: because the sensitivity of the acceleration sensor built in the earphone is high, the acquired acceleration data may include the body and the like, and the current motion state is not related to the current motion state.
  • the data of the abnormal jitter and the random noise such as the system measurement, the data pre-processing unit 121 first eliminates the abnormal jitter noise in the acceleration data through the 5 Hz low-pass filter, and then further filters the low-pass filter using the third-order moving average filter. Accurate acceleration data is obtained by random noise in the filtered acceleration data. Finally, accurate acceleration data is segmented through different time windows.
  • the segmented acceleration data contains acceleration characteristic information such as data distribution characteristics and peak-to-valley characteristics.
  • the feature extraction unit 122 is configured to extract acceleration feature information in different motion states according to the pre-processed acceleration data; the acceleration data distribution caused by different motion states is different, and the difference in motion intensity is reflected in the distribution center and dispersion degree of the acceleration data. Different, by extracting the acceleration data distribution characteristics under different motion states, combined with the peak-valley characteristics, to obtain different motion states
  • the acceleration characteristic information wherein the data distribution characteristics include an average value, an average difference, an interquartile range, a dispersion coefficient, a skewness coefficient, and the like.
  • the first state identifying unit 123 is configured to train the state recognition model in different motion states according to the extracted acceleration feature information, and initially identify the current motion state of the user by using the state recognition model, such as rest, walking, running, climbing, cycling, and the like.
  • the second state identifying unit 124 is configured to calculate a user heart rate value according to the user's heart rate, identify a typical motion state of the user according to the relationship between the user heart rate value interval and the motion state, and identify the current motion state of the user according to the state recognition model to the user.
  • Maximum heart rate therefore, the user's typical motion state can be initially identified according to the percentage of the maximum heart rate, as shown in Table 1 below:
  • the embodiment of the invention realizes very accurate user motion state recognition by combining user heart rate and acceleration data, and can distinguish the motion state in which the happiness rate data is relatively close, such as sleep and sit-in, and can also distinguish jogging, running, sprinting. Run and wait for different stages of exercise.
  • the audio play module 130 is configured to play music in a music playlist transmitted by the server.
  • the audio playback device further includes a setting module, and the setting module is configured to control a switch of the music switching function, and the user can close the function by setting the module when the music switching function is not needed, and the function is turned off. After that, the earphone stops detecting the user's heart rate and acceleration data, and the audio playback device will resume to manually switch the music.
  • the setting module can also set parameters such as the interval between switching music and the number of music of the music playlist transmitted by the server 300 according to personal needs, for example, 10 Minutes, half an hour, etc., to avoid switching music too often, or to play the same type of music in the same state of motion, affecting the user experience.
  • FIG. 2 is a schematic structural diagram of an audio playing system according to an embodiment of the present invention.
  • the audio playback system includes an earphone 200, an audio playback device 100 shown in FIG. 1, and a server 300, and the audio playback device 100 is connected to the headset 200 and the server 300, respectively.
  • the earphone 200 includes an earphone body (not shown), a heart rate sensor 210, an acceleration sensor 220, and a second communication module 230.
  • the heart rate sensor 210, the acceleration sensor 220, and the second communication module 230 are respectively disposed in the earphone body, and the heart rate is The sensor 210 and the acceleration sensor 220 are respectively connected to the second communication module 230 network.
  • the heart rate sensor 210 is a photoelectric heart rate sensor, and adopts an optical detection principle.
  • the incident light source uses a certain wavelength of light (such as 500 nm to 560 nm green light, 660 nm to 720 nm red light), transmits to the user at a certain angle, and then receives. After the optical signal reflected by the user, the optical signal reflected by the user is decomposed to obtain a PPG signal (electrocardiographic signal data), and the user's heart rate is detected according to the PPG signal.
  • a certain wavelength of light such as 500 nm to 560 nm green light, 660 nm to 720 nm red light
  • the acceleration sensor 220 is used to acquire acceleration data of the user during the wearing process, such as shaking, rising, falling, etc.; specifically, please refer to FIG. 3 and FIG. 4 together, FIG. 3 is a schematic diagram of the acceleration sensor acquiring data, and FIG. 4 is a walking process. Acceleration data obtained in . As shown in FIG. 3, the data acquired by the acceleration sensor 220 includes acceleration data in three directions of x, y, and z. The larger the positive value of the acceleration data in each direction, the faster the acceleration in the direction, and the acceleration in each direction. The smaller the negative value of the data, the faster the acceleration in the opposite direction.
  • the acquired PPG signal is relatively undulating, so the acceleration data of the earphone 200 is synchronously acquired by the acceleration sensor 220 for compensating the PPG obtained by the heart rate sensor 210. signal.
  • the second communication module 230 is configured to transmit the user heart rate acquired by the heart rate sensor 210 and the acceleration data acquired by the acceleration sensor 220 into the audio playback device 100 connected to the earphone 200.
  • the audio playback device 100 includes, but is not limited to, a smart phone, a tablet computer, a PC, etc.
  • the connection method of the earphone 200 and the audio playback device 100 includes but is not limited to wired (3.5 mm earphone jack), Bluetooth or WiFi, and the like.
  • connection mode if it is a wired connection mode, the communication mode of the second communication module 230 and the audio playback device 100 is: using the Audio line and the MIC line for communication interaction, if it is a Bluetooth or WiFi connection mode, the Bluetooth or WiFi communication interface can be directly used. Conduct communication interactions.
  • the earphone 200 of the embodiment of the present invention separately sets the heart rate sensor 210 and the acceleration sensor 220 on the basis of the hardware of the ordinary earphone 200.
  • the heart rate sensor 210 and the acceleration sensor 220 acquire the user's heart rate and acceleration data in real time. And transmitting the user heart rate and acceleration data to the audio playback device 100 connected to the earphone 200 in real time, so that the audio playback device 100 can play the matched music according to the user's heart rate and acceleration data without manual operation by the user; the earphone structure Simple, easy to develop, and expand the use of the ordinary headset 200.
  • a matching application software is installed in the audio playback device 100.
  • the application software can accurately identify the user's motion state according to the user's heart rate and acceleration data transmitted by the headset 200, and transmit the user heart rate and the user motion state to the server 300, and receive the transmission by the server 300.
  • a music playlist that matches the user's heart rate and user motion state, and plays the music in the music playlist.
  • the server 300 is configured to classify music according to parameters such as music type, language, motion state (sport type), and music rhythm, and select a music playlist matching the user's heart rate and user motion state from the music database, and play the music.
  • the list is transferred to the audio playback device for playback.
  • the server 300 includes a music classification module 310 and a music matching module 320;
  • the music classification module 310 is configured to classify music according to parameters such as music type, language, motion state, and music rhythm, and establish a corresponding music database according to the classification result; wherein the server 300 includes music of various types and various languages. First, they are classified according to the type of music and language, such as pop, rock, folk, electronic or European, American, Chinese, Japanese and Korean, and then classified according to the state of motion, such as sleeping, sitting, walking, jogging, etc.; and through the music rhythm detection algorithm From the PCM (Pulse Code Modulation) encoded music signal, the rhythm points and rhythm values of all the music in each motion state are detected, and the music in each motion state is further classified according to the rhythm points and the rhythm values of the music, Specifically, as shown in FIG. 4, it is a music database diagram of an embodiment of the present invention.
  • PCM Pulse Code Modulation
  • the music rhythm detection algorithm specifically includes: firstly, for the excitation detection, parsing out most of the signal excitation positions in the music from the input signal; secondly, detecting the rhythm, and estimating the possible rhythm values of the target music by using the obtained signal excitation position; Rhythm tracking, based on the output of the rhythm detection, derives the rhythm value of the music and identifies the specific location of the rhythm point.
  • the music matching module 320 is configured to receive a user heart rate and a user motion state transmitted by the audio playback device 100, match a corresponding music database according to the user motion state, and select a music playlist in which the music rhythm matches the user's heart rate in the music database, and The music playlist is transmitted to the audio playback device 100 for playing; in the embodiment of the present invention, the music playlist includes at least one piece of music, which can be specifically set by the setting module of the audio playback device 100; the music in the music playlist
  • the information only includes basic information such as song name, singer, etc.
  • the actual music information such as lyrics is still in the server 300, that is, the server 300 will match the user's motion state and the user's heart rate to the music to tell the audio player.
  • 100 is set to enable the audio playback device 100 to play the corresponding music without having to cache the entire piece of music in the audio playback device 100 to avoid occupying the storage space of the audio playback device 100.
  • the audio playback system of the embodiment of the present invention separately sets the heart rate sensor and the acceleration sensor on the basis of the hardware of the ordinary earphone, and in the process of wearing the earphone, the heart rate sensor and the acceleration sensor acquire the user heart rate and acceleration data in real time, and the user heart rate and The acceleration data is transmitted to the audio playback device connected to the earphone in real time, and the audio playback device accurately identifies the corresponding user motion state according to the user's heart rate and acceleration data, and transmits the user heart rate and the user motion state to the server, and the server transmits and the user's motion state.
  • the music playlists that match and the music rhythm match the user's heart rate are played, which increases the pleasure during the exercise and achieves better exercise without manual operation by the user.
  • the invention has less hardware, lower cost, simple design, convenient development, and expands the use function of the ordinary earphone, which is beneficial to improving the user experience and satisfying the individual needs of the user.
  • FIG. 5 is a flowchart of an audio playing method according to an embodiment of the present invention.
  • the audio playing method of the embodiment of the present invention includes the following steps:
  • Step 100 acquiring user heart rate and acceleration data in real time through the earphone
  • a heart rate sensor and an acceleration sensor are respectively disposed in the earphone, and the heart rate sensor is a photoelectric heart rate sensor.
  • the optical detection principle is adopted.
  • the incident light source adopts a certain wavelength of light, is transmitted to the user at a certain angle, and then receives the reflected by the user.
  • the returned optical signal is decomposed and decomposed by the optical signal reflected by the user to obtain a PPG signal, and the user's heart rate is detected by the PPG signal.
  • the acceleration sensor is used to acquire acceleration data of the user such as shaking, rising, falling, and the like. Acceleration sensor
  • the data obtained includes acceleration data in three directions of x, y, and z. The larger the positive value of the acceleration data in each direction, the faster the acceleration in the direction, and the smaller the negative value of the acceleration data in each direction indicates the opposite direction. The faster the acceleration.
  • Step 200 transmitting user heart rate and acceleration data to an audio playback device connected to the earphone;
  • connection or communication mode of the earphone with the audio playback device includes, but is not limited to, wired, Bluetooth, WiFi, and the like.
  • Step 300 The audio playback device receives user heart rate and acceleration data transmitted by the earphone, identifies a corresponding user motion state according to the user heart rate and acceleration data, and transmits the user heart rate and the user motion state to the server;
  • the audio playback device is installed with supporting application software, and the application software can accurately recognize the user's motion state according to the user's heart rate and acceleration data.
  • the method for identifying the motion state of the user specifically includes:
  • Step 301 Perform pre-processing on acceleration data transmitted by the earphone
  • the data pre-processing includes: due to the high sensitivity of the acceleration sensor built in the earphone, the acquired acceleration data may include abnormal noise such as abnormal vibration of the body and the current motion state, and random noise such as system measurement.
  • the abnormal jitter noise in the acceleration data is first eliminated by the 5 Hz low-pass filter, and then the third-order moving average filter is used to further filter the random noise in the acceleration data filtered by the low-pass filter to obtain an accurate Acceleration data; finally, the accurate acceleration data is segmented through different time windows, and the segmented acceleration data includes There are acceleration characteristic information such as data distribution characteristics and peak-valley characteristics.
  • Step 302 Extract acceleration characteristic information in different motion states according to the pre-processed acceleration data.
  • step 302 the acceleration data distribution caused by different motion states is different, and the difference in motion intensity is reflected in the distribution center and the dispersion degree of the acceleration data, by extracting the acceleration data distribution characteristics under different motion states, and combining the peaks and valleys. Value characteristics, thereby obtaining acceleration characteristic information under different motion states; wherein, the data distribution characteristics include an average value, an average difference, a quartile difference, a dispersion coefficient, a skewness coefficient, and the like.
  • Step 303 Train the state recognition model in different motion states according to the extracted acceleration feature information, and initially identify the current motion state of the user by using the state recognition model;
  • Step 304 Calculate a user heart rate value according to the user's heart rate, identify a typical motion state of the user according to the relationship between the user heart rate value interval and the motion state, and perform typical motion of the user according to the current motion state of the user initially identified by the state recognition model. The state is compensated to obtain the final user motion state;
  • the heart rate intervals corresponding to different motion states are different.
  • the user heart rate value interval is generally between 60 and 90, and the maximum heart rate of healthy adults is 220-age, and the exercise generally does not exceed the maximum. Heart rate, therefore, the user's typical motion state can be initially identified based on the percentage of the maximum heart rate. Since the contact distance between the heart rate sensor and the user's skin changes frequently during the exercise state, the heart rate of the user acquired by the earphone is relatively fluctuating, so the acceleration data of the earphone is synchronously acquired by the acceleration sensor to compensate the user's heart rate, thereby realizing very accurate user motion. State recognition, which can distinguish the movement state of the happiness rate data, such as sleep and sit-in, etc. To distinguish between different stages of exercise such as jogging, running, and sprinting.
  • Step 400 The server selects, from the music database, a music playlist that matches the user's motion state, and the music rhythm matches the user's heart rate, and transmits the music playlist to the audio playback device for playing;
  • the server includes music of various types and various languages, and the server first classifies according to the music type and language, such as pop, rock, folk, electronic or European, American, Chinese, Japanese, Korean, etc., and then according to the state of motion. Sorting, such as sleeping, sitting, walking, jogging, etc.; and detecting the rhythm points and rhythm values of all the music in each motion state from the PCM encoded music signal by the music rhythm detection algorithm, according to the rhythm points and rhythm values of the music The music in each sport state is further classified.
  • the music type and language such as pop, rock, folk, electronic or European, American, Chinese, Japanese, Korean, etc.
  • the music rhythm detection algorithm specifically includes: firstly, for the excitation detection, parsing out most of the signal excitation positions in the music from the input signal; secondly, detecting the rhythm, and estimating the possible rhythm values of the target music by using the obtained signal excitation position; Rhythm tracking, based on the output of the rhythm detection, derives the rhythm value of the music and identifies the specific location of the rhythm point.
  • the server After receiving the user's heart rate and the user's motion state transmitted by the audio playback device, the server matches the corresponding music database according to the user's motion state, selects a music playlist whose music rhythm matches the user's heart rate in the music database, and plays the music.
  • the list is transmitted to the audio playback device for playing.
  • the music playlist includes at least one piece of music, which can be specifically set by the audio playback device.
  • the audio playing method of the embodiment of the present invention sets the heart rate separately on the basis of the hardware of the ordinary earphone
  • the sensor and the acceleration sensor acquire the user's heart rate and acceleration data in real time through the heart rate sensor and the acceleration sensor during the wearing of the earphone, and transmit the user heart rate and acceleration data to the audio playback device connected to the earphone in real time, and the audio playback device is based on
  • the user heart rate and acceleration data accurately identify the corresponding user motion state, and transmit the user heart rate and the user motion state to the server, and the server transmits a music playlist that matches the user's motion state, and the music rhythm matches the user's heart rate, and increases Pleasure during exercise for better exercise without the need for manual operation by the user.
  • the invention has less hardware, lower cost, simple design, convenient development, and expands the use function of the ordinary earphone, which is beneficial to improving the user experience and satisfying the individual needs of the user.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Physiology (AREA)
  • Reverberation, Karaoke And Other Acoustics (AREA)
  • Electrophonic Musical Instruments (AREA)
  • Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)

Abstract

一种音频播放装置(100)、系统及方法，涉及音频播放技术领域。所述音频播放装置(100)用于接收耳机(200)传输的用户心率和加速度数据，根据所述用户心率和加速度数据识别出用户运动状态，将所述用户心率和用户运动状态传输至服务器(300)，接收服务器(300)传输的与所述用户心率及用户运动状态相匹配的音乐播放列表，并播放所述音乐播放列表中的音乐。该装置可播放与用户心率及用户运动状态相匹配的音乐，从而增加运动过程中的愉悦感，达到更好的运动效果，无需用户手动操作，有利于提升用户的使用体验，满足用户的个性化需求。

Description

一种音频播放装置、系统及方法 【技术领域】
本发明实施例涉及音频播放技术领域，尤其涉及一种音频播放装置、系统及方法。
【背景技术】
随着生活水平的不断提高，人们越来越重视健康而会选择各种运动，例如走路、跑步、骑车、爬山等。不过运动本身是一件挺枯燥的事情，尤其是一个人在健身房或操场上跑步，需要一些能够帮助人们缓解枯燥的方法，而听音乐是大多数人的一个选择。有了音乐的陪伴不仅会让人感到心情愉悦，此外激昂的音乐还会激发人们运动的潜能和培养坚强的意志，从而达到更好的运动效果。
但是不同的运动场景需要与合适的音乐相配合，如果选择节奏感和人的心跳速率相吻合的音乐，人会随着音乐的节奏而运动起来，可以使运动更轻松，也更容易坚持；但是如果选择不合适的音乐，例如快跑时播放古典音乐，反而不利于步伐和呼吸的控制，而更容易受伤。
目前主流的音频播放装置，都无法针对不同的运动状态自动选择或切换播放的音乐类型，也没有根据不同运动状态所匹配的音乐节奏进行细分的音乐库，用户需要手动选择音乐类别，或者播放指定的歌曲。如果能够自动识别用户运动状态，并自动播放或切换与用户当前运动状态相吻合、节奏和心率一致的音乐，无疑会大大提高用户的使用感受，满足用户的个性化需求。
【发明内容】
本发明实施例提供了一种音频播放装置、系统及方法，旨在至少在一定程度上解决现有技术中的上述技术问题之一。
为了解决以上提出的问题，本发明实施例采用的技术方案包括：
一种音频播放装置，所述音频播放装置用于接收耳机传输的用户心率和加速度数据，根据所述用户心率和加速度数据识别出用户运动状态，将所述用户心率和用户运动状态传输至服务器，接收服务器传输的与所述用户心率及用户运动状态相匹配的音乐播放列表，并播放所述音乐播放列表中的音乐。
本发明实施例采取的技术方案还包括：所述音频播放装置包括：
第一通信模块：用于所述音频播放装置与耳机及服务器之间的通信交互；
状态识别模块：用于根据用户心率和加速度数据识别出对应的用户运动状态；
音频播放模块：用于播放与所述服务器传输的音乐播放列表相对应的音乐。
本发明实施例采取的技术方案还包括：所述状态识别模块包括：
数据预处理单元：用于对所述耳机传输的加速度数据进行预处理；
特征提取单元：用于根据预处理后的加速度数据提取不同运动状态下的加速度特征信息；
第一状态识别单元：用于根据提取的加速度特征信息训练不同运动状态下的状态识别模型，通过状态识别模型初步识别出用户的当前运动状态；
第二状态识别单元：用于根据用户心率计算用户心率值，根据用户心率值区间与运动状态的关系识别出用户的典型运动状态，并根据所述状态识别模型初步识别出的用户的当前运动状态对所述用户的典型运动状态进行补偿，得到最终的用户运动状态。
本发明实施例采取的技术方案还包括：所述数据预处理单元对耳机传输的加速度数据进行预处理具体为：通过低通滤波器过滤所述耳机传输的加速度数据中的噪音，然后使用三阶滑动平均滤波器进一步滤除经低通滤波器过滤后的加速度数据中的随机噪音，得到准确的加速度数据；并通过不同的时间窗口对所述准确的加速度数据进行分割处理。
本发明实施例采取的另一技术方案为：一种音频播放系统，包括音频播放装置以及耳机和服务器；
所述耳机用于获取用户心率和加速度数据，并将获取到的用户心率和加速度数据传输至音频播放装置；
所述音频播放装置用于根据所述用户心率和加速度数据识别出用户运动状态，并将所述用户心率和用户运动状态传输至服务器；
所述服务器用于根据所述用户运动状态匹配对应的音乐数据库，在匹配的音乐数据库中选择音乐节奏与所述用户心率相匹配的音乐播放列表，并将所述音乐播放列表传输至音频播放装置进行播放。
本发明实施例采取的技术方案还包括：所述耳机包括心率传感器，所述心率传感器用于获取用户心率。
本发明实施例采取的技术方案还包括：所述耳机还包括加速度传感器，所述加速度传感器用于获取用户的加速度数据；所述加速度传感器获取的加速度数据包括x、y、z三个方向上的数据。
本发明实施例采取的技术方案还包括：所述耳机还包括第二通信模块，所述第二通信模块用于将所述用户心率和加速度数据传输至与所述耳机连接的音频播放装置。
本发明实施例采取的技术方案还包括：所述服务器包括：
音乐分类模块：用于根据音乐类型、语种、运动状态及音乐节奏对音乐进行分类，并根据分类结果建立对应的音乐数据库；
音乐匹配模块：用于接收音频播放装置传输的用户心率及用户运动状态，根据用户运动状态匹配对应的音乐数据库，在匹配的音乐数据库中选择音乐节奏与用户心率相吻合的音乐播放列表，并将所述音乐播放列表传输至音频播放装置。
本发明实施例采取的技术方案还包括：所述音乐分类模块根据音乐节奏对音乐进行分类的分类方式为：通过音乐节奏检测算法从PCM编码音乐信号中检测音乐的节奏点以及节奏值，根据音乐的节奏点以及节奏值对每种运动状态下的音乐进行分类。
本发明实施例采取的又一技术方案：一种音频播放方法，包括以下步骤：
步骤a：接收耳机传输的用户心率和加速度数据；
步骤b：根据所述用户心率和加速度数据识别出用户运动状态，将所述用户 心率和用户运动状态传输至服务器；
步骤c：接收服务器传输的与所述用户心率及用户运动状态相匹配的音乐播放列表，并播放所述音乐播放列表中的音乐。
本发明实施例采取的技术方案还包括：在所述步骤a中，所述耳机中分别设有心率传感器和加速度传感器，所述心率传感器用于获取用户心率；所述加速度传感器用于获取x、y、z三个方向上的加速度数据。
本发明实施例采取的技术方案还包括：在所述步骤b中，所述根据所述用户心率和加速度数据识别出用户运动状态的识别方式包括：
步骤b1：对所述耳机传输的加速度数据进行预处理；
步骤b2：根据预处理后的加速度数据提取不同运动状态下的加速度特征信息；
步骤b3：根据提取的加速度特征信息训练不同运动状态下的状态识别模型，通过状态识别模型初步识别出用户的当前运动状态；
步骤b4：根据用户心率计算用户心率值，根据用户心率值区间与运动状态的关系识别出用户的典型运动状态，并根据所述状态识别模型初步识别出的用户的当前运动状态对所述用户的典型运动状态进行补偿，得到最终的用户运动状态。
本发明实施例采取的技术方案还包括：在所述步骤b1中，所述数据预处理具体包括：通过低通滤波器过滤所述耳机传输的加速度数据中的噪音，然后使用三阶滑动平均滤波器进一步滤除经低通滤波器过滤后的加速度数据中的随机 噪音，得到准确的加速度数据；并通过不同的时间窗口对所述准确的加速度数据进行分割处理。
本发明实施例采取的技术方案还包括：所述步骤c还包括：通过所述服务器根据音乐类型、语种、运动状态及音乐节奏对音乐进行分类，并根据分类结果建立对应的音乐数据库。
本发明实施例采取的技术方案还包括：所述根据音乐节奏对音乐进行分类的分类方式为：通过音乐节奏检测算法从PCM编码音乐信号中检测音乐的节奏点以及节奏值，根据音乐的节奏点以及节奏值对每种运动状态下的音乐进行分类。
与现有技术相比，本发明实施例的有益效果在于：本发明实施例的音频播放装置、系统及方法通过耳机内的心率传感器和加速度传感器实时获取用户心率和加速度数据，并将用户心率和加速度数据传输至与该耳机连接的音频播放装置上，音频播放装置根据用户心率和加速度数据精确识别对应的用户运动状态；并将用户心率和用户运动状态传输至服务器，服务器传输与用户运动状态相匹配，且音乐节奏与用户心率相吻合的音乐播放列表进行播放。本发明可以增加运动过程中的愉悦感，达到更好的运动效果，而无需用户手动操作。相对于现有技术，本发明增加的硬件较少，成本较低，设计简单，开发便捷，扩充了普通耳机的使用功能，有利于提升用户的使用体验，满足用户的个性化需求。
【附图说明】
图1是本发明实施例的音频播放装置的结构示意图；
图2是本发明实施例的音频播放系统的结构示意图；
图3是加速度传感器获取数据示意图；
图4是走路过程中获取到的加速度数据；
图5是本发明实施例的音频播放方法的流程图。
【具体实施方式】
为了便于理解本发明，下面将参照相关附图对本发明进行更全面的描述。附图中给出了本发明的较佳实施例。但是，本发明可以以许多不同的形式来实现，并不限于本文所描述的实施例。相反地，提供这些实施例的目的是使对本发明的公开内容的理解更加透彻全面。
除非另有定义，本文所使用的所有的技术和科学术语与属于本发明的技术领域的技术人员通常理解的含义相同。本文中在本发明的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本发明。
请参阅图1，是本发明实施例的音频播放装置的结构示意图。本发明实施例的音频播放装置100中安装有应用软件，应用软件可以根据耳机传输的用户心率和加速度数据精确识别出用户运动状态，将用户心率和用户运动状态传输至服务器，接收服务器传输的与用户心率及用户运动状态相匹配的音乐播放列表，并播放音乐播放列表中的音乐。具体地，音频播放装置100包括第一通信模块110、状态识别模块120和音频播放模块130；
第一通信模块110用于音频播放装置100与耳机及服务器之间的通信交互； 具体包括：接收耳机传输的用户心率和加速度数据、将用户心率及状态识别模块120识别到的用户运动状态传输至服务器300、接收服务器传输的与用户心率及用户运动状态相匹配的音乐播放列表；第一通信模块110的通信方式包括但不限于有线、蓝牙或WiFi等。
状态识别模块120用于根据用户心率和加速度数据识别出对应的用户运动状态，并将用户心率和用户运动状态由第一通信模块110传输至服务器300。具体地，状态识别模块120包括数据预处理单元121、特征提取单元122、第一状态识别单元123和第二状态识别单元124；
数据预处理单元121用于对耳机获取的加速度数据进行预处理；预处理具体包括：由于耳机内设的加速度传感器的灵敏度较高，其获取到的加速度数据中可能包含身体等与当前运动状态无关的不正常抖动的噪音和系统测量等随机噪音，数据预处理单元121首先通过5Hz低通滤波器消除加速度数据中的不正常抖动噪音，然后使用三阶滑动平均滤波器进一步滤除经低通滤波器过滤后的加速度数据中的随机噪音，得到准确的加速度数据；最后通过不同的时间窗口将准确的加速度数据进行分割，分割后的加速度数据含有数据分布特征和峰谷值特征等加速度特征信息。
特征提取单元122用于根据预处理后的加速度数据提取不同运动状态下的加速度特征信息；不同的运动状态引起的加速度数据分布情况不同，同时运动的强度差异体现在加速度数据的分布中心和离散程度不同，通过提取不同运动状态下的加速度数据分布特征，并结合峰谷值特征，从而得到不同运动状态下 的加速度特征信息；其中，数据分布特征包括平均值、平均差、四分位差、离散系数、偏态系数等。
第一状态识别单元123用于根据提取的加速度特征信息训练不同运动状态下的状态识别模型，通过状态识别模型初步识别出用户的当前运动状态，例如静止、走路、跑步，爬山，骑车等。
第二状态识别单元124用于根据用户心率计算用户心率值，根据用户心率值区间与运动状态的关系识别出用户的典型运动状态，并根据状态识别模型初步识别出的用户的当前运动状态对用户的典型运动状态进行补偿，得到最终的用户运动状态；其中，在静止状态下，用户心率值一般在60～90之间，健康成年人的最大心率＝220-年龄，运动时一般不会超过该最大心率，因此，可以根据最大心率的百分比对用户的典型运动状态进行初步的识别，具体如下表1所示：
表1心率值和运动状态的关系
Figure PCTCN2016108971-appb-000001
Figure PCTCN2016108971-appb-000002
本发明实施例通过用户心率和加速度数据相结合的方式实现非常精确的用户运动状态识别，可以区分开心率数据较为接近的运动状态，例如睡眠和静坐等，也可以区分开慢跑、快跑、冲刺跑等不同阶段的运动状态。
音频播放模块130用于播放服务器传输的音乐播放列表中的音乐。
在本发明实施例中，音频播放装置还包括一设置模块，设置模块用于控制音乐切换功能的开关，用户可以在不需要使用音乐切换功能时，通过设置模块对该功能进行关闭，关闭该功能后，耳机停止检测用户心率和加速度数据，音频播放装置将恢复为手动切换播放音乐；设置模块还可以根据个人需求设置切换音乐的间隔时间、服务器300传输音乐播放列表的音乐数量等参数，例如10分钟、半小时等，避免过于频繁的切换音乐，或在同一运动状态下循环播放同一类型的音乐，影响用户的使用体验。
请参阅图2，是本发明实施例的音频播放系统的结构示意图。本发明实施例 的音频播放系统包括耳机200、图1所示的音频播放装置100和服务器300，音频播放装置100分别与耳机200和服务器300网络连接。
具体地，耳机200包括耳机主体(图未示)、心率传感器210、加速度传感器220和第二通信模块230；心率传感器210、加速度传感器220和第二通信模块230分别设置于耳机主体内，且心率传感器210和加速度传感器220分别与第二通信模块230网络连接。
其中，心率传感器210为光电式心率传感器，采用光学检测原理，入射光源采用一定波长的光(如500nm-560nm的绿光，660nm-720nm的红光)，以一定的角度透射至用户，然后接收经过用户反射回来的光信号，并对用户反射回来的光信号进行分解后获取PPG信号(心电信号数据)，根据PPG信号检测用户心率。
加速度传感器220用于在佩戴过程中获取用户的加速度数据，例如晃动、上升、下降等；具体地，请一并参阅图3和图4，图3是加速度传感器获取数据示意图，图4是走路过程中获取到的加速度数据。如图3所示，加速度传感器220获取的数据包括x、y、z三个方向上的加速度数据，各个方向上的加速度数据正值越大说明该方向上的加速度越快，各个方向上的加速度数据负值越小说明反方向的加速度越快。由于在运动状态下心率传感器210与用户皮肤的接触距离时常会发生变化，导致获取的PPG信号起伏比较大，因此通过加速度传感器220同步获取耳机200的加速度数据，用于补偿心率传感器210获得的PPG信号。
第二通信模块230用于将心率传感器210获取的用户心率和加速度传感器220获取的加速度数据传输到与该耳机200连接的音频播放装置100中。在本发明实施例中，音频播放装置100包括但不限于智能手机、平板电脑或PC等，耳机200与音频播放装置100的连接方法包括但不限于有线(3.5mm耳机孔)、蓝牙或WiFi等连接方式，如果是有线连接方式，第二通信模块230与音频播放装置100的通信方式为：使用Audio线和MIC线进行通信交互，如果是蓝牙或WiFi连接方式，可以直接使用蓝牙或WiFi通信接口进行通信交互。
本发明实施例的耳机200通过在普通耳机200的硬件基础上分别设置心率传感器210和加速度传感器220，在佩戴耳机200的过程中，通过心率传感器210和加速度传感器220实时获取用户心率和加速度数据，并将用户心率和加速度数据实时传输至与该耳机200连接的音频播放装置100上，使得音频播放装置100可以根据用户心率和加速度数据播放与之相匹配的音乐，无需用户手动操作；该耳机结构简单，开发便捷，扩充了普通耳机200的使用功能。
音频播放装置100中安装有配套的应用软件，应用软件可以根据耳机200传输的用户心率和加速度数据精确识别出用户运动状态，将用户心率和用户运动状态传输至服务器300，接收服务器300传输的与用户心率及用户运动状态相匹配的音乐播放列表，并播放音乐播放列表中的音乐。
服务器300用于根据音乐类型、语种、运动状态(运动类型)及音乐节奏等参数对音乐进行分类，并从音乐数据库中选择与用户心率及用户运动状态相匹配的音乐播放列表，将该音乐播放列表传输至音频播放装置进行播放。
具体地，服务器300包括音乐分类模块310和音乐匹配模块320；
音乐分类模块310用于根据音乐类型、语种、运动状态及音乐节奏等参数对音乐进行分类，并根据分类结果建立对应的音乐数据库；其中，服务器300中包括各种类型、各种语种的音乐，首先分别根据音乐类型和语种进行分类，例如流行、摇滚、民谣、电子或欧美、华语、日韩等，然后根据运动状态进行分类，例如睡觉、静坐、走路、慢跑等；并通过音乐节奏检测算法从PCM(脉冲编码调制，Pulse Code Modulation)编码音乐信号中检测出每种运动状态下所有音乐的节奏点以及节奏值，根据音乐的节奏点以及节奏值对每种运动状态下的音乐进一步分类，具体如图4所示，是本发明实施例的音乐数据库图。
音乐节奏检测算法具体包括：首先为激励检测，从输入信号中解析出音乐中大部分的信号激励位置；其次为节奏检测，用得到的信号激励位置估计目标音乐可能的几个节奏值；最后为节奏跟踪，基于节奏检测的输出，得出该音乐的节奏值，并标识出节奏点的具体位置。
音乐匹配模块320用于接收音频播放装置100传输的用户心率及用户运动状态，根据用户运动状态匹配对应的音乐数据库，在该音乐数据库中选择音乐节奏与用户心率相吻合的音乐播放列表，并将该音乐播放列表传输至音频播放装置100进行播放；在本发明实施例中，音乐播放列表中至少包括一首音乐，具体可通过音频播放装置100的设置模块进行设定；音乐播放列表中的音乐信息仅包括歌曲名称、歌手等基本信息，歌词等实际音乐信息依然是在服务器300中，即服务器300将与用户运动状态和用户心率匹配的是音乐告诉音频播放装 置100，以使音频播放装置100播放对应的音乐，而无需将整首音乐全部缓存在音频播放装置100中，避免占有音频播放装置100的存储空间。
本发明实施例的音频播放系统通过在普通耳机的硬件基础上分别设置心率传感器和加速度传感器，在佩戴耳机的过程中，通过心率传感器和加速度传感器实时获取用户心率和加速度数据，并将用户心率和加速度数据实时传输至与该耳机连接的音频播放装置上，音频播放装置根据用户心率和加速度数据精确识别对应的用户运动状态，并将用户心率和用户运动状态传输至服务器，服务器传输与用户运动状态相匹配，且音乐节奏与用户心率相吻合的音乐播放列表进行播放，增加运动过程中的愉悦感，达到更好的运动效果，而无需用户手动操作。相对于现有技术，本发明增加的硬件较少，成本较低，设计简单，开发便捷，扩充了普通耳机的使用功能，有利于提升用户的使用体验，满足用户的个性化需求。
请参阅图5，是本发明实施例的音频播放方法的流程图。本发明实施例的音频播放方法包括以下步骤：
步骤100：通过耳机实时获取用户心率和加速度数据；
在步骤100中，耳机中分别设有心率传感器和加速度传感器，心率传感器为光电式心率传感器，采用光学检测原理，入射光源采用一定波长的光，以一定的角度透射至用户，然后接收经过用户反射回来的光信号，并对用户反射回来的光信号进行分解后获取PPG信号，通过PPG信号检测用户心率。加速度传感器用于获取用户的加速度数据例如晃动、上升、下降等。加速度传感器获 取的数据包括x、y、z三个方向上的加速度数据，各个方向上的加速度数据正值越大说明该方向上的加速度越快，各个方向上的加速度数据负值越小说明反方向的加速度越快。
步骤200：将用户心率和加速度数据传输到与该耳机连接的音频播放装置中；
在步骤200中，耳机与音频播放装置的连接或通信方式包括但不限于有线、蓝牙或WiFi等。
步骤300：音频播放装置接收耳机传输的用户心率和加速度数据，根据用户心率和加速度数据识别出对应的用户运动状态，并将用户心率和用户运动状态传输至服务器；
在步骤300中，音频播放装置中安装有配套的应用软件，应用软件根据用户心率和加速度数据可以精确识别出用户运动状态。用户运动状态的识别方法具体包括：
步骤301：对耳机传输的加速度数据进行预处理；
在步骤301中，数据预处理包括：由于耳机内设的加速度传感器的灵敏度较高，其获取到的加速度数据中可能包含身体等与当前运动状态无关的不正常抖动的噪音和系统测量等随机噪音，本发明实施例首先通过5Hz低通滤波器消除加速度数据中的不正常抖动噪音，然后使用三阶滑动平均滤波器进一步滤除经低通滤波器过滤后的加速度数据中的随机噪音，得到准确的加速度数据；最后通过不同的时间窗口将准确的加速度数据进行分割，分割后的加速度数据含 有数据分布特征和峰谷值特征等加速度特征信息。
步骤302：根据预处理后的加速度数据提取不同运动状态下的加速度特征信息；
在步骤302中，不同的运动状态引起的加速度数据分布情况不同，同时运动的强度差异体现在加速度数据的分布中心和离散程度不同，通过提取不同运动状态下的加速度数据分布特征，并结合峰谷值特征，从而得到不同运动状态下的加速度特征信息；其中，数据分布特征包括平均值、平均差、四分位差、离散系数、偏态系数等。
步骤303：根据提取的加速度特征信息训练不同运动状态下的状态识别模型，通过状态识别模型初步识别出用户的当前运动状态；
步骤304：根据用户心率计算用户心率值，根据用户心率值区间与运动状态的关系识别出用户的典型运动状态，并根据状态识别模型初步识别出的用户的当前运动状态对所述用户的典型运动状态进行补偿，得到最终的用户运动状态；
在步骤304中，不同运动状态对应的心率区间不同，在静止状态下，用户心率值区间一般在60～90之间，健康成年人的最大心率＝220-年龄，运动时一般不会超过该最大心率，因此，可以根据最大心率的百分比对用户典型运动状态进行初步的识别。由于在运动状态下心率传感器与用户皮肤的接触距离时常会发生变化，导致耳机获取的用户心率起伏比较大，因此通过加速度传感器同步获取耳机的加速度数据对用户心率进行补偿，实现非常精确的用户运动状态识别，可以区分开心率数据较为接近的运动状态，例如睡眠和静坐等，也可 以区分开慢跑、快跑、冲刺跑等不同阶段的运动状态。
步骤400：服务器从音乐数据库中选择与用户运动状态相匹配，且音乐节奏与用户心率相吻合的音乐播放列表，并将该音乐播放列表传输至音频播放装置进行播放；
在步骤400中，服务器中包括各种类型、各种语种的音乐，服务器首先分别根据音乐类型和语种进行分类，例如流行、摇滚、民谣、电子或欧美、华语、日韩等，然后根据运动状态进行分类，例如睡觉、静坐、走路、慢跑等；并通过音乐节奏检测算法从PCM编码音乐信号中检测出每种运动状态下所有音乐的节奏点以及节奏值，根据音乐的节奏点以及节奏值对每种运动状态下的音乐进一步分类。
音乐节奏检测算法具体包括：首先为激励检测，从输入信号中解析出音乐中大部分的信号激励位置；其次为节奏检测，用得到的信号激励位置估计目标音乐可能的几个节奏值；最后为节奏跟踪，基于节奏检测的输出，得出该音乐的节奏值，并标识出节奏点的具体位置。
服务器在接收到音频播放装置传输的用户心率及用户运动状态后，根据用户运动状态匹配对应的音乐数据库，在该音乐数据库中选择音乐节奏与用户心率相吻合的音乐播放列表，并将该音乐播放列表传输至音频播放装置进行播放；在本发明实施例中，音乐播放列表中至少包括一首音乐，具体可通过音频播放装置进行设定。
本发明实施例的音频播放方法通过在普通耳机的硬件基础上分别设置心率 传感器和加速度传感器，在佩戴耳机的过程中，通过心率传感器和加速度传感器实时获取用户心率和加速度数据，并将用户心率和加速度数据实时传输至与该耳机连接的音频播放装置上，音频播放装置根据用户心率和加速度数据精确识别对应的用户运动状态，并将用户心率和用户运动状态传输至服务器，服务器传输与用户运动状态相匹配，且音乐节奏与用户心率相吻合的音乐播放列表进行播放，增加运动过程中的愉悦感，达到更好的运动效果，而无需用户手动操作。相对于现有技术，本发明增加的硬件较少，成本较低，设计简单，开发便捷，扩充了普通耳机的使用功能，有利于提升用户的使用体验，满足用户的个性化需求。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到各实施方式可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件。基于这样的理解，上述技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存储介质中，如ROM/RAM、磁碟、光盘等，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行各个实施例或者实施例的某些部分所述的方法。
上述实施例为本发明较佳的实施方式，但本发明的实施方式并不受上述实施例的限制，其他的任何未背离本发明的精神实质与原理下所作的改变、修饰、替代、组合、简化，均应为等效的置换方式，都包含在本发明的保护范围之内。

Claims (16)

  1. 一种音频播放装置，其特征在于，所述音频播放装置用于接收耳机传输的用户心率和加速度数据，根据所述用户心率和加速度数据识别出用户运动状态，将所述用户心率和用户运动状态传输至服务器，接收服务器传输的与所述用户心率及用户运动状态相匹配的音乐播放列表，并播放所述音乐播放列表中的音乐。
  2. 根据权利要求1所述的音频播放装置，其特征在于，所述音频播放装置包括：
    第一通信模块：用于所述音频播放装置与耳机及服务器之间的通信交互；
    状态识别模块：用于根据用户心率和加速度数据识别出对应的用户运动状态；
    音频播放模块：用于播放与所述服务器传输的音乐播放列表相对应的音乐。
  3. 根据权利要求2所述的音频播放装置，其特征在于，所述状态识别模块包括：
    数据预处理单元：用于对所述耳机传输的加速度数据进行预处理；
    特征提取单元：用于根据预处理后的加速度数据提取不同运动状态下的加速度特征信息；
    第一状态识别单元：用于根据提取的加速度特征信息训练不同运动状态下的状态识别模型，通过状态识别模型初步识别出用户的当前运动状态；
    第二状态识别单元：用于根据用户心率计算用户心率值，根据用户心率值区间与运动状态的关系识别出用户的典型运动状态，并根据所述状态识别模型初步识别出的用户的当前运动状态对所述用户的典型运动状态进行补偿，得到最终的用户运动状态。
  4. 根据权利要求3所述的音频播放装置，其特征在于，所述数据预处理单元对耳机传输的加速度数据进行预处理具体为：通过低通滤波器过滤所述耳机传输的加速度数据中的噪音，然后使用三阶滑动平均滤波器进一步滤除经低通 滤波器过滤后的加速度数据中的随机噪音，以得到准确的加速度数据；并通过不同的时间窗口对所述准确的加速度数据进行分割处理。
  5. 一种音频播放系统，其特征在于，包括权利要求1至4任一项所述的音频播放装置，以及耳机和服务器；
    所述耳机用于获取用户心率和加速度数据，并将获取到的用户心率和加速度数据传输至音频播放装置；
    所述音频播放装置用于根据所述用户心率和加速度数据识别出用户运动状态，并将所述用户心率和用户运动状态传输至服务器；
    所述服务器用于根据所述用户运动状态匹配对应的音乐数据库，在匹配的音乐数据库中选择音乐节奏与所述用户心率相匹配的音乐播放列表，并将所述音乐播放列表传输至音频播放装置进行播放。
  6. 根据权利要求5所述的音频播放系统，其特征在于，所述耳机包括心率传感器，所述心率传感器用于获取用户心率。
  7. 根据权利要求6所述的音频播放系统，其特征在于，所述耳机还包括加速度传感器，所述加速度传感器用于获取用户的加速度数据；所述加速度传感器获取的加速度数据包括x、y、z三个方向上的数据。
  8. 根据权利要求7所述的音频播放系统，其特征在于，所述耳机还包括第二通信模块，所述第二通信模块用于将所述用户心率和加速度数据传输至与所述耳机连接的音频播放装置。
  9. 根据权利要求5所述的音频播放系统，其特征在于，所述服务器包括：
    音乐分类模块：用于根据音乐类型、语种、运动状态及音乐节奏对音乐进行分类，并根据分类结果建立对应的音乐数据库；
    音乐匹配模块：用于接收音频播放装置传输的用户心率及用户运动状态，根据用户运动状态匹配对应的音乐数据库，在匹配的音乐数据库中选择音乐节奏与用户心率相吻合的音乐播放列表，并将所述音乐播放列表传输至音频播放装置。
  10. 根据权利要求9所述的音频播放系统，其特征在于，所述音乐分类模块根据音乐节奏对音乐进行分类的分类方式为：通过音乐节奏检测算法从PCM编码音乐信号中检测音乐的节奏点以及节奏值，根据音乐的节奏点以及节奏值对每种运动状态下的音乐进行分类。
  11. 一种音频播放方法，其特征在于，包括以下步骤：
    步骤a：接收耳机传输的用户心率和加速度数据；
    步骤b：根据所述用户心率和加速度数据识别出用户运动状态，将所述用户心率和用户运动状态传输至服务器；
    步骤c：接收服务器传输的与所述用户心率及用户运动状态相匹配的音乐播放列表，并播放所述音乐播放列表中的音乐。
  12. 根据权利要求11所述的音频播放方法，其特征在于，在所述步骤a中，所述耳机中分别设有心率传感器和加速度传感器，所述心率传感器用于获取用户心率；所述加速度传感器用于获取x、y、z三个方向上的加速度数据。
  13. 根据权利要求12所述的音频播放方法，其特征在于，在所述步骤b中，所述根据所述用户心率和加速度数据识别出用户运动状态的识别方式包括：
    步骤b1：对所述耳机传输的加速度数据进行预处理；
    步骤b2：根据预处理后的加速度数据提取不同运动状态下的加速度特征信息；
    步骤b3：根据提取的加速度特征信息训练不同运动状态下的状态识别模型，通过状态识别模型初步识别出用户的当前运动状态；
    步骤b4：根据用户心率计算用户心率值，根据用户心率值区间与运动状态的关系识别出用户的典型运动状态，并根据所述状态识别模型初步识别出的用户的当前运动状态对所述用户的典型运动状态进行补偿，得到最终的用户运动状态。
  14. 根据权利要求13所述的音频播放方法，其特征在于，在所述步骤b1中，所述数据预处理具体包括：通过低通滤波器过滤所述耳机传输的加速度数 据中的噪音，然后使用三阶滑动平均滤波器进一步滤除经低通滤波器过滤后的加速度数据中的随机噪音，以得到准确的加速度数据；并通过不同的时间窗口对所述准确的加速度数据进行分割处理。
  15. 根据权利要求11所述的音频播放方法，其特征在于，所述步骤c还包括：通过所述服务器根据音乐类型、语种、运动状态及音乐节奏对音乐进行分类，并根据分类结果建立对应的音乐数据库。
  16. 根据权利要求15所述的音频播放方法，其特征在于，所述根据音乐节奏对音乐进行分类的分类方式为：通过音乐节奏检测算法从PCM编码音乐信号中检测音乐的节奏点以及节奏值，根据音乐的节奏点以及节奏值对每种运动状态下的音乐进行分类。
PCT/CN2016/108971 2016-12-08 2016-12-08 一种音频播放装置、系统及方法 WO2018103028A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201680001779.6A CN106687958A (zh) 2016-12-08 2016-12-08 一种音频播放装置、系统及方法
PCT/CN2016/108971 WO2018103028A1 (zh) 2016-12-08 2016-12-08 一种音频播放装置、系统及方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2016/108971 WO2018103028A1 (zh) 2016-12-08 2016-12-08 一种音频播放装置、系统及方法

Publications (1)

Publication Number Publication Date
WO2018103028A1 true WO2018103028A1 (zh) 2018-06-14

Family

ID=58849592

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/108971 WO2018103028A1 (zh) 2016-12-08 2016-12-08 一种音频播放装置、系统及方法

Country Status (2)

Country Link
CN (1) CN106687958A (zh)
WO (1) WO2018103028A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110347864A (zh) * 2019-07-18 2019-10-18 广东小天才科技有限公司 一种智能调节音频参数的方法及系统

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107301871A (zh) * 2017-05-25 2017-10-27 武汉云信众汇通讯科技有限公司 一种自适应音乐播放方法
CN108989706A (zh) * 2017-06-02 2018-12-11 北京字节跳动网络技术有限公司 基于音乐节奏生成特效的方法及装置
CN107016996B (zh) * 2017-06-06 2020-11-10 广东小天才科技有限公司 一种音频数据的处理方法及装置
CN107594739A (zh) * 2017-09-03 2018-01-19 泉州迪特工业产品设计有限公司 一种智能炫酷鞋及其实现方法
CN107689229A (zh) * 2017-09-25 2018-02-13 广东小天才科技有限公司 一种用于可穿戴设备的语音处理方法及装置
CN108089833B (zh) * 2017-11-01 2021-04-30 捷开通讯(深圳)有限公司 智能移动终端及其播放音乐的方法、具有存储功能的装置
CN108491539A (zh) * 2018-04-02 2018-09-04 Oppo广东移动通信有限公司 终端控制方法、装置、存储介质和电子设备
CN108429970B (zh) * 2018-05-28 2020-04-14 Oppo广东移动通信有限公司 音频播放方法、装置、终端、耳机及可读存储介质
CN108429972B (zh) * 2018-05-28 2019-12-27 Oppo广东移动通信有限公司 音乐播放方法、装置、终端、耳机及可读存储介质
CN111260864A (zh) * 2018-11-30 2020-06-09 珠海格力电器股份有限公司 语音播报的控制方法、装置和空调
CN110841244A (zh) * 2019-10-29 2020-02-28 刘孝章 一种人体肢体运动和器械运动模拟声音增强运动能力的装置及其运行方法
CN111202509B (zh) * 2020-01-17 2023-07-04 山东中医药大学 基于听觉表现策略的靶心率监控方法及装置
CN111739530A (zh) * 2020-06-05 2020-10-02 北京搜狗科技发展有限公司 一种交互方法、装置、耳机和耳机收纳装置
CN111831116A (zh) * 2020-07-01 2020-10-27 深圳大学 一种基于ppg信息的智能设备交互方法
CN112765395B (zh) * 2021-01-22 2023-09-19 咪咕音乐有限公司 音频播放方法、电子设备和存储介质
CN114305373A (zh) * 2021-12-31 2022-04-12 佳禾智能科技股份有限公司 运动心率监测耳机

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140277649A1 (en) * 2013-03-15 2014-09-18 Futurewei Technologies, Inc. Music Recommendation Based on Biometric and Motion Sensors on Mobile Device
CN104133551A (zh) * 2014-07-15 2014-11-05 惠州Tcl移动通信有限公司 根据心率值改变歌曲的音乐风格的方法和移动终端
CN104822103A (zh) * 2015-03-26 2015-08-05 北京国承万通信息科技有限公司 耳机
CN105744420A (zh) * 2016-03-23 2016-07-06 惠州Tcl移动通信有限公司 智能运动耳机及智能运动系统
CN105810218A (zh) * 2014-12-31 2016-07-27 深圳富泰宏精密工业有限公司 根据运动状态自动选歌的系统及方法
CN106128487A (zh) * 2016-07-05 2016-11-16 乐视控股（北京）有限公司 音频文件播放方法及装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103839538B (zh) * 2012-11-22 2016-01-20 腾讯科技（深圳）有限公司 音乐节奏检测方法及检测装置
CN104361016B (zh) * 2014-10-15 2018-05-29 广东小天才科技有限公司 一种根据运动状态调整音乐播放效果的方法及装置
CN105139875B (zh) * 2015-09-24 2017-07-11 湖北文理学院 基于运动时海拔高度变化的音频播放方法及装置
CN105832320A (zh) * 2016-02-01 2016-08-10 乐视移动智能信息技术（北京）有限公司 一种根据用户的运动状态自动选择音乐进行播放的方法和系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140277649A1 (en) * 2013-03-15 2014-09-18 Futurewei Technologies, Inc. Music Recommendation Based on Biometric and Motion Sensors on Mobile Device
CN104133551A (zh) * 2014-07-15 2014-11-05 惠州Tcl移动通信有限公司 根据心率值改变歌曲的音乐风格的方法和移动终端
CN105810218A (zh) * 2014-12-31 2016-07-27 深圳富泰宏精密工业有限公司 根据运动状态自动选歌的系统及方法
CN104822103A (zh) * 2015-03-26 2015-08-05 北京国承万通信息科技有限公司 耳机
CN105744420A (zh) * 2016-03-23 2016-07-06 惠州Tcl移动通信有限公司 智能运动耳机及智能运动系统
CN106128487A (zh) * 2016-07-05 2016-11-16 乐视控股（北京）有限公司 音频文件播放方法及装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110347864A (zh) * 2019-07-18 2019-10-18 广东小天才科技有限公司 一种智能调节音频参数的方法及系统

Also Published As

Publication number Publication date
CN106687958A (zh) 2017-05-17

Similar Documents

Publication Publication Date Title
WO2018103028A1 (zh) 一种音频播放装置、系统及方法
Schuller et al. The interspeech 2018 computational paralinguistics challenge: Atypical & self-assessed affect, crying & heart beats
CN107427716B (zh) 人类绩效优化与训练的方法及系统
US20200012682A1 (en) Biometric-music interaction methods and systems
Mannini et al. Activity recognition in youth using single accelerometer placed at wrist or ankle
CN103561652B (zh) 用于辅助患者的方法和系统
JP6812604B2 (ja) オーディオアクティビティ追跡および要約
KR101749706B1 (ko) 사용자 장치를 사용하여 획득된 상황 정보 및 생체 정보를 기반으로 사용자의 감정을 예측하는 방법 및 시스템
US11205408B2 (en) Method and system for musical communication
CN104036776A (zh) 一种应用于移动终端的语音情感识别方法
Ellamil et al. One in the dance: musical correlates of group synchrony in a real-world club environment
Abushakra et al. Acoustic signal classification of breathing movements to virtually aid breath regulation
JP2014526057A (ja) 音声分析の方法およびシステム
CN110706816B (zh) 一种基于人工智能进行睡眠环境调控的方法及设备
US11253747B2 (en) Automated activity detection and tracking
KR102015097B1 (ko) 생체정보를 이용한 감성 검출 장치 및 감성 검출 방법을 실행하기 위한 프로그램을 기록한 컴퓨터로 판독 가능한 기록매체
Niu et al. Research on genetic algorithm based on emotion recognition using physiological signals
CN113448438B (zh) 一种基于睡眠感知的控制系统及方法
CN106599057A (zh) 一种自适应用户情绪的音乐节奏调控终端及其方法
Nathan et al. EMOSIC—An emotion based music player for Android
CN114569132A (zh) 一种基于示教的婴儿训练系统
Fan et al. A new approach for smoking event detection using a variational autoencoder and neural decision forest
CN108475295A (zh) 用于预测即将进食时刻的可穿戴系统
CN107463646A (zh) 一种助眠音乐智能推荐方法及装置
Luo et al. Design of embedded real-time system for snoring and OSA detection based on machine learning

Legal Events

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

Ref document number: 16923282

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16923282

Country of ref document: EP

Kind code of ref document: A1