CN108245171B - Method for obtaining parameter model, fatigue detection method and device, medium and equipment - Google Patents
Method for obtaining parameter model, fatigue detection method and device, medium and equipment Download PDFInfo
- Publication number
- CN108245171B CN108245171B CN201711466152.0A CN201711466152A CN108245171B CN 108245171 B CN108245171 B CN 108245171B CN 201711466152 A CN201711466152 A CN 201711466152A CN 108245171 B CN108245171 B CN 108245171B
- Authority
- CN
- China
- Prior art keywords
- unit time
- blink
- parameter model
- current unit
- blink parameter
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
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
- A61B5/1103—Detecting eye twinkling
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Physiology (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Evolutionary Computation (AREA)
- Signal Processing (AREA)
- Psychiatry (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Fuzzy Systems (AREA)
- Ophthalmology & Optometry (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The present disclosure relates to a method for obtaining a parameter model, a fatigue detection method and apparatus, a medium, and a device, the method comprising: calculating the correlation coefficient of the sampling point data in each target unit time in the N target unit times and the sampling point data in the current unit time one by one; inputting M groups of sampling point data with the maximum correlation with the sampling point data in the current unit time into an input layer of a blink parameter prediction model to predict the blink parameter model; and finally determining the blink parameter model output from the output layer of the blink parameter prediction model as the blink parameter model in the current unit time. Therefore, the accuracy of the obtained blink parameter model can be improved, the accurate blink parameter model can be obtained based on the interference signal or the incomplete signal by using the method, the application range of the method is effectively widened, the accuracy of the result of fatigue detection based on the blink parameter model can be ensured, and the use experience of a user is improved.
Description
Technical Field
The disclosure relates to the field of eye blink detection, in particular to a method for acquiring a parameter model, a fatigue detection method and device, a medium and equipment.
Background
The body movement detecting chip integrates a whole set of circuits with electromagnetic wave transmitting and receiving functions into one chip for detecting the movement of a human body, and has the advantages of small volume, low power consumption and convenient use. The basic principle is to emit electromagnetic waves to the outside, detect the reflected electromagnetic waves and output the detection result in a voltage mode. When the body movement detection chip works, the electromagnetic wave transmitting and receiving are a continuous process, and the output voltage is a continuously changing process. When the human body action is not detected, the output voltage is stabilized in a numerical range with extremely tiny change; when human body action is detected, the output voltage fluctuates, the amplitude of the fluctuation corresponds to the amplitude of the detected human body action, and the frequency of the fluctuation corresponds to the frequency of the human body action. The detection result output by the body movement detection chip is a path of analog signal with continuously changed voltage, and the analog signal is output to the outside in a chip pin mode.
In a complicated environment, it is difficult for the body motion detection chip to output a complete signal under the influence of a noise signal, and it is difficult to obtain a complete signal even after the signal is subjected to noise reduction. A blink detection result obtained based on the signal of the disability is inaccurate, and even the blink detection is difficult according to the signal of the disability, so that it is difficult to detect whether the user is in a fatigue state based on the blink detection result.
Disclosure of Invention
In order to solve the above problems, the present disclosure provides a method of obtaining a parameter model, a fatigue detection method and apparatus, a medium, and a device.
To achieve the above object, according to a first aspect of the present disclosure, there is provided a method of acquiring a blink parameter model, the method comprising:
calculating correlation coefficients of sampling point data in each target unit time in N target unit times and sampling point data in the current unit time one by one, wherein the N target unit times are N continuous unit times before the current unit time, and N is a positive integer greater than or equal to 2;
inputting M groups of sampling point data with the maximum correlation with the sampling point data in the current unit time into an input layer of a blink parameter prediction model to predict the blink parameter model, wherein the input layer of the blink parameter prediction model comprises M input neurons, the M groups of sampling point data correspond to the M input neurons one by one, and for each input neuron, a correlation coefficient between the sampling point data corresponding to the input neuron and the sampling point data in the current unit time is used as a weight between the input neuron and a target intermediate neuron, wherein the target intermediate neuron is any neuron in a first-level hidden layer of the blink parameter prediction model, and M is more than or equal to 2 and less than or equal to N;
and finally determining the blink parameter model output from the output layer of the blink parameter prediction model as the blink parameter model in the current unit time.
Optionally, the N target unit times include a unit time previous to the current unit time.
According to a second aspect of the present disclosure, there is provided a fatigue detection method, the method comprising:
calculating a blink parameter model in the current unit time according to the obtained sampling point data in the current unit time, wherein the blink parameter model in the unit time at least comprises blink time;
when the blink parameter model in the current unit time does not meet a first preset condition, reserving the calculated blink parameter model in the current unit time; when the blink parameter model in the current unit time meets the first preset condition, ignoring the computed blink parameter model in the current unit time, and re-determining the blink parameter model in the current unit time by using the method of the first aspect, wherein the first preset condition is that: calculating that the deviation between the blink parameter model in the current unit time and an ideal blink parameter model in the unit time exceeds an acceptable level, wherein any parameter in the blink parameter model in the current unit time exceeds a parameter range formed by the parameter in the blink parameter models in the N target unit times;
and determining whether the detected user is in a fatigue state or not at least according to the blink parameter model in the current unit time.
Optionally, the determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time includes:
taking the next unit time of the current unit time as a new current unit time, returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, and determining that the detected user is in a fatigue state, wherein the second preset condition is as follows: when P continuous unit time periods exist, the blink time is larger than or equal to the blink time threshold, and P is a positive integer larger than or equal to 2.
Optionally, the determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time includes:
taking the next unit time of the current unit time as a new current unit time, and returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, wherein the second preset condition is as follows: when the blink time in the blink parameter model in P continuous unit time is larger than or equal to the blink time threshold value, P is a positive integer larger than or equal to 2;
determining a blink parameter model in each unit time by using the method of the first aspect one by one from the next unit time of the P unit times;
and when no blink parameter model with the blink time smaller than the blink time threshold value exists in Q unit times from the next unit time of the P unit times, determining that the detected user is in the fatigue state, wherein Q is a positive integer larger than or equal to 2.
Optionally, the determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time further includes:
when a blink parameter model with blink time smaller than the blink time threshold exists in the Q unit time, determining that the detected user is not in a fatigue state;
the method further comprises the following steps:
and taking the next unit time of the unit time corresponding to the blink parameter model with the blink time smaller than the blink time threshold value in the Q unit times as new current unit time, and returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time.
According to a third aspect of the present disclosure, there is provided an apparatus for acquiring a blink parameter model, the apparatus comprising:
the first calculation module is used for calculating correlation coefficients of sampling point data in each target unit time in N target unit times and sampling point data in the current unit time one by one, wherein the N target unit times are N continuous unit times before the current unit time, and N is a positive integer greater than or equal to 2;
the prediction module is used for inputting M groups of sampling point data with the maximum correlation with the sampling point data in the current unit time into an input layer of a blink parameter prediction model to predict the blink parameter model, wherein the input layer of the blink parameter prediction model comprises M input neurons, the M groups of sampling point data correspond to the M input neurons one by one, and for each input neuron, a correlation coefficient between the sampling point data corresponding to the input neuron and the sampling point data in the current unit time is used as a weight between the input neuron and a target intermediate neuron, wherein the target intermediate neuron is any neuron in a first-level hidden layer of the blink parameter prediction model, and M is more than or equal to 2 and less than or equal to N;
a first determining module, configured to finally determine the blink parameter model output from the output layer of the blink parameter prediction model as the blink parameter model in the current unit time.
Optionally, the N target unit times include a unit time previous to the current unit time.
According to a fourth aspect of the present disclosure, there is provided a fatigue detection apparatus, the apparatus comprising:
the second calculation module is used for calculating a blink parameter model in the current unit time according to the obtained sampling point data in the current unit time, wherein the blink parameter model in the unit time at least comprises blink time;
the second determining module is used for reserving the calculated blink parameter model in the current unit time when the blink parameter model in the current unit time does not meet a first preset condition; when the blink parameter model in the current unit time meets the first preset condition, ignoring the blink parameter model in the current unit time calculated by the second calculating module, and re-determining the blink parameter model in the current unit time by using the apparatus of the second aspect, wherein the first preset condition is that: the second calculation module calculates that the deviation between the blink parameter model in the current unit time and the ideal blink parameter model in the unit time exceeds an acceptable level, and any parameter in the blink parameter model in the current unit time exceeds a parameter range formed by the parameter in the blink parameter models in the N target unit times;
and the third determining module is used for determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time.
Optionally, the third determining module is configured to:
taking the next unit time of the current unit time as a new current unit time, triggering the second calculating module to calculate a blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, and determining that the detected user is in a fatigue state, wherein the second preset condition is as follows: when P continuous unit time periods exist, the blink time is larger than or equal to the blink time threshold, and P is a positive integer larger than or equal to 2.
Optionally, the third determining module includes:
the first determining submodule is used for taking the next unit time of the current unit time as a new current unit time, triggering the second calculating module to calculate the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, wherein the second preset condition is as follows: when the blink time in the blink parameter model in P continuous unit time is larger than or equal to the blink time threshold value, P is a positive integer larger than or equal to 2;
a second determining submodule, configured to determine, one by one, a blink parameter model in each unit time from a next unit time of the P unit times, using the apparatus of the second aspect;
and the third determining submodule is used for determining that the detected user is in a fatigue state when no blink parameter model with the blink time smaller than the blink time threshold exists in Q unit time from the next unit time of the P unit time, wherein Q is a positive integer larger than or equal to 2.
Optionally, the third determining module further includes:
a fourth determining submodule, configured to determine that the detected user is not in a fatigue state when a blink parameter model with a blink time smaller than the blink time threshold exists in the Q unit times;
the device further comprises:
and the fourth determining module is used for taking the next unit time of the unit time corresponding to the blink parameter model with the blink time smaller than the blink time threshold value in the Q unit times as the new current unit time, and triggering the second calculating module to calculate the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time.
According to a fifth aspect of the present disclosure, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method of the first aspect.
According to a sixth aspect of the present disclosure, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method of the second aspect.
According to a seventh aspect of the present disclosure, there is provided an electronic apparatus comprising:
the computer-readable storage medium of the fifth aspect; and
one or more processors to execute the program in the computer-readable storage medium.
According to an eighth aspect of the present disclosure, there is provided an electronic apparatus comprising:
the computer-readable storage medium of the sixth aspect; and
one or more processors to execute the program in the computer-readable storage medium.
By the technical scheme, on one hand, the blink parameter model in the current unit time can be effectively and accurately predicted according to the sampling point data with high relevance to the current unit time, and the accuracy of the obtained blink parameter model can be improved. On the other hand, when the signal output by the body motion detection chip is interfered or has a defect, the accurate blink parameter model can be obtained based on the interference signal or the defect signal by using the method, the application range of the method is effectively widened, the accuracy of the result of fatigue detection based on the blink parameter model can be ensured, and the use experience of a user is improved.
Additional features and advantages of the disclosure will be set forth in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure without limiting the disclosure. In the drawings:
fig. 1 is a flow chart of a method of obtaining a blink parameter model provided according to an embodiment of the disclosure;
FIG. 2 is a flow chart of a fatigue detection method provided according to an embodiment of the present disclosure;
fig. 3 is a block diagram of an apparatus for acquiring a blink parameter model provided according to an embodiment of the disclosure;
FIG. 4 is a block diagram of a fatigue detection apparatus provided in accordance with an embodiment of the present disclosure;
FIG. 5 is a block diagram illustrating an electronic device in accordance with an exemplary embodiment;
FIG. 6 is a block diagram illustrating an electronic device in accordance with an example embodiment.
Detailed Description
The following detailed description of specific embodiments of the present disclosure is provided in connection with the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the present disclosure, are given by way of illustration and explanation only, not limitation.
Fig. 1 is a flowchart illustrating a method for acquiring a blink parameter model according to an embodiment of the disclosure, where the method includes:
in S11, correlation coefficients of the sampling point data in each of N target unit times and the sampling point data in the current unit time are calculated one by one, where the N target unit times are N unit times that are consecutive before the current unit time, and N is a positive integer greater than or equal to 2.
Alternatively, the correlation coefficient of the sample point data in each target unit time and the sample point data in the current unit time may be calculated as follows:
wherein r represents a correlation coefficient between the sampling point data in the target unit time and the sampling point data in the current unit time;
n represents the number of sampling points, wherein the number of the sampling points in the target unit time is the same as the number of the sampling points in the current unit time;
x represents a voltage value of a sampling point in a target unit time;
y represents a voltage value of a sampling point in the current unit time;
In S12, inputting M sets of sample point data in the target unit time having the highest correlation with the sample point data in the current unit time into an input layer of a blink parameter prediction model for predicting the blink parameter model, wherein the input layer of the blink parameter prediction model includes M input neurons, the M sets of sample point data are in one-to-one correspondence with the M input neurons, and for each input neuron, a correlation coefficient between the sample point data corresponding to the input neuron and the sample point data in the current unit time is used as a weight between the input neuron and a target interneuron, wherein the target interneuron is any neuron in a first-level hidden layer of the blink parameter prediction model, and 2 ≦ M ≦ N.
Wherein the blink parameter prediction model is pre-constructed according to the data of the sampling points in a large amount of unit time and the blink parameter model. The blink parameter prediction model can be constructed through a neural network algorithm, and the neural network algorithm is the prior art and is not described herein again. And taking a correlation coefficient of the sampling point data corresponding to the input neuron and the sampling point data in the current unit time as the weight between the input neuron and the target intermediate neuron, and training the correlation weight in the blink parameter prediction model according to the correlation coefficient so as to improve the accuracy of the output blink parameter model of the blink parameter prediction model.
In S13, the blink parameter model output from the output layer of the blink parameter prediction model is finally determined as the blink parameter model in the current unit time.
By the technical scheme, on one hand, the blink parameter model in the current unit time can be effectively and accurately predicted according to the sampling point data with high relevance to the current unit time, and the accuracy of the obtained blink parameter model can be improved. On the other hand, when the signal output by the body motion detection chip is interfered or has a defect, the accurate blink parameter model can be obtained based on the interference signal or the defect signal by using the method, the application range of the method is effectively widened, and effective data support is provided for relevant detection based on the blink parameter model.
Optionally, the N target unit times include a unit time previous to the current unit time; the sampling point data in each unit time with the largest correlation with the sampling point data in the current unit time is generally the sampling point data in each unit time with the closest distance to the current unit time, so that the accuracy of the blink parameter model output by the blink parameter prediction model can be further improved by taking the sampling point data in N unit times adjacent to the current unit time as a reference for performing correlation calculation on the sampling point data in the current unit time.
Based on the method provided by the disclosure, an accurate blink parameter model corresponding to the signal can be obtained based on the signal output by the body motion detection chip, and when the signal output by the body motion detection chip is interfered or incomplete, the accurate blink parameter model can be obtained based on the method, so that fatigue detection can be performed according to the signal output by the body motion detection chip by using the method. Specifically, as shown in fig. 2, a flowchart of a fatigue detection method according to an embodiment of the present disclosure is provided, and as shown in fig. 2, the method includes:
in S21, a blink parameter model in the current unit time is calculated according to the obtained sampling point data in the current unit time, wherein the blink parameter model in the unit time at least includes the blink time.
For example, the blink parameter model for the current unit time may be calculated as follows:
and recording instant sampling point data Temp (v, t) in unit time, wherein t is sampling time, and v is a voltage value output by the body motion detection chip at the time t. When v (temp) continuously increases along with T, recording the time point when v (temp) starts to increase as T _ begin, recording the point when v (temp) starts to decrease as T (peak), and recording the time value of the sampling point before the first sampling point when v (temp) starts to increase after T (peak) as T _ end. Wherein t (m) represents a time value corresponding to m, and v (m) represents a voltage value corresponding to m.
The blink time T _ bli is T _ end-T _ begin. And if a plurality of blink time periods are determined in the current unit time, determining the average value of the blink time periods as the blink time in the blink parameter model in the current unit time.
Optionally, the blink parameter model may further include the following parameters:
v (peak): maximum value of voltage value of sampling point;
n: the number of blinks;
t _ int: a blink time interval;
tr: the time required for the voltage value of the sampling point to increase from v (T _ begin) to v (peak);
tf: the time required for the voltage value of the sampling point to decrease from v (peak) to v (T _ end).
In step S22, when the blink parameter model in the current unit time does not satisfy the first preset condition, the calculated blink parameter model in the current unit time is retained; when the blink parameter model in the current unit time meets the first preset condition, ignoring the computed blink parameter model in the current unit time, and re-determining the blink parameter model in the current unit time by using the method for acquiring the blink parameter model provided by the disclosure, wherein the first preset condition is as follows: the calculated deviation between the blink parameter model in the current unit time and the ideal blink parameter model in the unit time exceeds an acceptable level, and any parameter in the blink parameter model in the current unit time exceeds a parameter range formed by the parameter in the blink parameter models in the N target unit times.
In an embodiment, the calculated deviation between the blink parameter model for the current unit time and the ideal blink parameter model for the unit time exceeding the acceptable level may be:
the deviation of any parameter in the blink parameter model in the current unit time and the parameter in the ideal blink parameter model in the unit time exceeds the acceptable deviation range corresponding to the parameter. Illustratively, the blink parameter model includes 5 parameters, each of which has a corresponding deviation acceptable range, for example, the deviation acceptable range corresponding to the blink time is [ -a, a ], the deviation between the blink time in the blink parameter model in the current unit time and the blink time in the ideal blink parameter model in the unit time is b, when b < -a or b > a, the deviation between the blink time in the blink parameter model in the current unit time and the blink time in the ideal blink parameter model in the unit time exceeds the deviation acceptable range corresponding to the blink time, so that the deviation between the blink parameter model in the current unit time and the ideal blink parameter model in the unit time exceeds an acceptable level.
In another embodiment, the calculated deviation between the blink parameter model for the current unit time and the ideal blink parameter model for the unit time exceeding an acceptable level may be:
the deviation between the parameters in the blink parameter model in the current unit time and the corresponding parameters in the ideal blink parameter model in the unit time exceeds the deviation acceptable range corresponding to the parameters, and the proportion of the number of the parameters with the deviation exceeding the deviation acceptable range to the total number of the parameters in the blink parameter model reaches a preset proportion. For example, the preset proportion is B, the blink parameter model includes s parameters in total, where there are t parameters whose deviation from the corresponding parameter in the ideal blink parameter model in the unit time exceeds the acceptable deviation range, and when (t/s) ≧ B, it may be determined that the deviation between the blink parameter model in the current unit time and the ideal blink parameter model in the unit time exceeds an acceptable level.
And forming a parameter range corresponding to each parameter according to the numerical values of various parameters in the blink parameter model in the N target unit times. For example, taking the blink time as an example, the minimum value Tmin of the blink time in the blink parameter model for N target unit times may be taken as the minimum value of the parameter range, the maximum value Tmax of the blink time in the blink parameter model for N target unit times may be taken as the maximum value of the parameter range, and then the parameter range formed by the blink time in the blink parameter model for N target unit times is [ Tmin, Tmax ]. And if the blink time T in the blink parameter model in the current unit time meets T < Tmin or T > Tmax, determining that the blink time exceeds a parameter range formed by the blink times in the blink parameter model in the N target unit times.
Therefore, when the blink parameter model in the current unit time meets the first preset condition, it indicates that at least one parameter in the blink parameter model in the current unit time is abnormal, that is, the calculated blink parameter model in the current unit time is inaccurate. At this time, the blink parameter model in the current unit time can be obtained according to the method for obtaining the blink parameter model provided by the disclosure, so that the accuracy of the blink parameter model in the current unit time can be ensured, and the influence on the fatigue detection result due to the inaccuracy of the blink parameter model can be avoided. Meanwhile, when the blink parameter model in the current unit time does not meet the first preset condition, the blink parameter model is accurate, can be directly applied to subsequent fatigue detection, can avoid repeated acquisition of the blink parameter model, saves time and improves processing efficiency.
At S23, it is determined whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time.
Optionally, an example implementation manner of determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time is as follows, including:
taking the next unit time of the current unit time as a new current unit time, returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, and determining that the detected user is in a fatigue state, wherein the second preset condition is as follows: when P continuous unit time periods exist, the blink time is larger than or equal to the blink time threshold, and P is a positive integer larger than or equal to 2.
After the blink parameter model in the current unit time is obtained, when the blink time in the blink parameter model is determined to be larger than or equal to the blink time threshold value, the number of unit times of which the blink time in the continuous blink parameter model is larger than or equal to the blink time threshold value is recorded. Illustratively, the number may be recorded by a counter, the counter is initialized to 0, and the accumulation operation is performed when it is determined that the blink time in the blink parameter model in the current unit time is greater than or equal to the blink time threshold; when the blink time in the blink parameter model in the current unit time is smaller than the blink time threshold, an initialization operation is performed, where the counting manner is only an example implementation manner, and is not limited in this disclosure.
Illustratively, T1-T7 are 7 consecutive units of time, P is 5, and the blink time threshold is 1 s. And when the blink parameter model of T1 is obtained, determining whether the blink time in the blink parameter model is larger than or equal to 1S, if the blink time in the blink parameter model is smaller than 1S, returning to S12 to calculate the blink parameter model of T2, and when the blink time in the blink parameter model of T2 is determined to be larger than or equal to 1S, performing accumulation operation by a counter. Then, the determined blink time in the blink parameter models of T3, T4 and T5 is greater than or equal to 1s, and the value of the counter is 4. And then, determining whether the blink time in the blink parameter model of T6 is greater than or equal to 1s, if the blink time in the blink parameter model of T6 is greater than or equal to 1s, performing accumulation operation on a counter, wherein the value of the counter is 5 at the moment, and the fatigue state of the detected user can be determined when a second preset condition is met. If the blinking time in the blinking parameter model of T6 is less than 1s, the counter is initialized, and the value of the counter is 0, and then the above steps are repeated from T7 to perform fatigue detection, which is not described herein again.
Wherein, when the human is in a normal state, the blinking time of the human is generally within a normal range, such as 0.2s-0.4 s. Therefore, when the blink time of the detected user in the blink parameter model in the continuous unit times is larger than or equal to the blink time threshold value, the blink time of the detected user in the continuous unit times is longer, and at the moment, the detected user can be determined to be in a fatigue state.
In the technical scheme, the signals output by the body movement detection chip are continuously monitored, so that the state of the detected user can be detected according to the blink parameter model of the user in a plurality of unit times, whether the detected user is in a fatigue state can be timely and accurately determined, and the detected user can be conveniently and correspondingly adjusted according to the state of the detected user.
Optionally, another example implementation manner of determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time is as follows, including:
taking the next unit time of the current unit time as a new current unit time, and returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, wherein the second preset condition is as follows: when the blink time in the blink parameter model in P continuous unit time is larger than or equal to the blink time threshold value, P is a positive integer larger than or equal to 2;
determining the blink parameter model in each unit time by utilizing the method for acquiring the blink parameter model provided by the disclosure one by one from the next unit time of the P unit times;
and when no blink parameter model with the blink time smaller than the blink time threshold value exists in Q unit times from the next unit time of the P unit times, determining that the detected user is in the fatigue state, wherein Q is a positive integer larger than or equal to 2.
The method for determining whether the blink parameter model in the current unit time meets the second preset condition is the same as the above, and is not described herein again. Illustratively, the current unit time is T10, and the blink parameter model for the current unit time satisfies a second preset condition.
For a unit time from the unit time T11, a blink parameter model for each unit time is determined using the method for acquiring a blink parameter model provided by the present disclosure.
Illustratively, Q takes a value of 5 and the blink time threshold is 1 s. In one embodiment, if the blinking time in the blinking parameter model in the unit time of T11-T15 is greater than or equal to 1s, the tested user may be determined to be in a fatigue state.
In the technical scheme, when the blink time in the blink parameter model in a plurality of continuous unit time is larger than or equal to the blink time threshold value, the phenomenon that the blink time of the detected user is too long continuously is shown. Then, the blink parameter model in the unit time from the unit time can be quickly and accurately obtained directly through the method for obtaining the blink parameter model provided by the disclosure. Therefore, when the detected user continuously has the phenomenon of long blinking time, the blinking parameter model in the unit time after the detected user is continuously detected, the influence of accidental conditions on fatigue detection can be effectively avoided, and the accuracy of fatigue detection is further improved.
Optionally, after the detected user is determined to be in the fatigue state, the detected user can be prompted so as to prompt the user in time, so that the dangerous situation that the user is in the fatigue state possibly can be effectively avoided, and the use requirements of the user can be met.
Optionally, the determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time further includes:
and when a blink parameter model with the blink time smaller than the blink time threshold exists in the Q unit time, determining that the detected user is not in a fatigue state.
The method further comprises the following steps:
and taking the next unit time of the unit time corresponding to the blink parameter model with the blink time smaller than the blink time threshold value in the Q unit times as new current unit time, and returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time.
In another embodiment, if the blink time in the blink parameter model of T12 is 0.8s, that is, the blink time is less than the blink time threshold value of 1s, it may be determined that the detected user is not in a fatigue state. And taking T13 as a new current unit time, and returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time, namely calculating the blink parameter model of T13.
When the blinking time of the detected user in the plurality of continuous unit times is larger than or equal to the blinking time threshold value, the blinking time of the detected user in the plurality of continuous unit times is longer. When the blink time in a plurality of subsequent unit times is not continuously kept to be larger than or equal to the blink time threshold value, the detected user is not in a fatigue state at the moment, and the next unit time of the unit time corresponding to the blink parameter model with the blink time smaller than the blink time threshold value can be used as the new current unit time to continuously detect the state of the detected user. By the technical scheme, the accuracy of the fatigue detection result can be effectively improved, and the state of the detected user can be monitored in real time.
In summary, in the above technical solution, when the blink parameter model in the current unit time is obtained according to the sampling point data in the current unit time, it is determined whether the blink parameter model is the blink parameter model in the current unit time, so that an influence of inaccuracy of the blink parameter model on the subsequent fatigue detection can be avoided. When the blink parameter model in the current unit time meets the first preset condition, namely the blink parameter model is not the blink parameter model in the current unit time, the blink parameter model in the current unit time is obtained through the method for obtaining the blink parameter model. The blink parameter model in the current unit time can be obtained through the method for obtaining the blink parameter model, and the accuracy of the blink parameter model is improved, so that accurate data support is provided for subsequent fatigue detection. Through the technical scheme, on one hand, the accuracy of the blink parameter model in the current unit time can be ensured, so that the accuracy of the result of fatigue detection based on the blink parameter model can be ensured. On the other hand, the fatigue detection can be carried out based on the incomplete signal output by the body movement detection chip, the application range of the method is expanded, and the use experience of a user is improved.
The present disclosure also provides an apparatus for acquiring a blink parameter model, as shown in fig. 3, where the apparatus 10 includes:
a first calculating module 100, configured to calculate correlation coefficients of sampling point data in each target unit time of N target unit times and sampling point data in a current unit time one by one, where the N target unit times are N continuous unit times before the current unit time, and N is a positive integer greater than or equal to 2;
a prediction module 200, configured to input M sets of sample point data having the highest correlation with the sample point data in the current unit time into an input layer of a blink parameter prediction model to perform prediction of the blink parameter model, where the input layer of the blink parameter prediction model includes M input neurons, the M sets of sample point data correspond to the M input neurons one to one, and for each input neuron, a correlation coefficient between the sample point data corresponding to the input neuron and the sample point data in the current unit time is used as a weight between the input neuron and a target intermediate neuron, where the target intermediate neuron is any neuron in a first-level hidden layer of the blink parameter prediction model, and M is greater than or equal to 2 and less than or equal to N;
a first determining module 300, configured to finally determine the blink parameter model output from the output layer of the blink parameter prediction model as the blink parameter model in the current unit time.
Optionally, the N target unit times include a unit time previous to the current unit time.
The present disclosure also provides a fatigue detecting device, as shown in fig. 4, the device 20 includes:
a second calculating module 400, configured to calculate a blink parameter model in the current unit time according to the obtained sample point data in the current unit time, where the blink parameter model in the unit time at least includes blink time;
a second determining module 500, configured to, when the blink parameter model in the current unit time does not meet a first preset condition, reserve the calculated blink parameter model in the current unit time; when the blink parameter model in the current unit time meets the first preset condition, ignoring the blink parameter model in the current unit time calculated by the second calculation module, and re-determining the blink parameter model in the current unit time by using the apparatus for acquiring the blink parameter model provided by the present disclosure, wherein the first preset condition is that: the second calculation module calculates that the deviation between the blink parameter model in the current unit time and the ideal blink parameter model in the unit time exceeds an acceptable level, and any parameter in the blink parameter model in the current unit time exceeds a parameter range formed by the parameter in the blink parameter models in the N target unit times;
a third determining module 600, configured to determine whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time.
Optionally, the third determining module 600 is configured to:
taking the next unit time of the current unit time as a new current unit time, triggering the second calculating module to calculate a blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, and determining that the detected user is in a fatigue state, wherein the second preset condition is as follows: when P continuous unit time periods exist, the blink time is larger than or equal to the blink time threshold, and P is a positive integer larger than or equal to 2.
Optionally, the third determining module 600 includes:
the first determining submodule is used for taking the next unit time of the current unit time as a new current unit time, triggering the second calculating module to calculate the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, wherein the second preset condition is as follows: when the blink time in the blink parameter model in P continuous unit time is larger than or equal to the blink time threshold value, P is a positive integer larger than or equal to 2;
the second determining submodule is used for determining the blink parameter model in each unit time by utilizing the device for acquiring the blink parameter model provided by the disclosure one by one from the next unit time of the P unit times;
and the third determining submodule is used for determining that the detected user is in a fatigue state when no blink parameter model with the blink time smaller than the blink time threshold exists in Q unit time from the next unit time of the P unit time, wherein Q is a positive integer larger than or equal to 2.
Optionally, the third determining module further includes:
a fourth determining submodule, configured to determine that the detected user is not in a fatigue state when a blink parameter model with a blink time smaller than the blink time threshold exists in the Q unit times;
the apparatus 20 further comprises:
and the fourth determining module is used for taking the next unit time of the unit time corresponding to the blink parameter model with the blink time smaller than the blink time threshold value in the Q unit times as the new current unit time, and triggering the second calculating module to calculate the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time.
Fig. 5 is a block diagram illustrating an electronic device 700 according to an example embodiment. As shown in fig. 5, the electronic device 700 may include: a processor 701, a memory 702, multimedia components 703, input/output (I/O) interfaces 704, and communication components 705.
The processor 701 is configured to control the overall operation of the electronic device 700 to complete all or part of the steps of the method for acquiring the blink parameter model or the fatigue detection method. The memory 702 is used to store various types of data to support operation at the electronic device 700, such as instructions for any application or method operating on the electronic device 700 and application-related data, such as contact data, transmitted and received messages, pictures, audio, video, and the like. The Memory 702 may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk, or optical disk. The multimedia components 703 may include screen and audio components. Wherein the screen may be, for example, a touch screen and the audio component is used for outputting and/or inputting audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may further be stored in the memory 702 or transmitted through the communication component 705. The audio assembly also includes at least one speaker for outputting audio signals. The I/O interface 704 provides an interface between the processor 701 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 705 is used for wired or wireless communication between the electronic device 700 and other devices. Wireless Communication, such as Wi-Fi, bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination of one or more of them, so that the corresponding Communication component 705 may include: Wi-Fi module, bluetooth module, NFC module.
In an exemplary embodiment, the electronic Device 700 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components for performing the above-described method for obtaining a blink parameter model or fatigue detection method.
In another exemplary embodiment, a computer readable storage medium comprising program instructions, such as the memory 702 comprising program instructions, executable by the processor 701 of the electronic device 700 to perform the method of obtaining a blink parameter model or the fatigue detection method described above is also provided.
Fig. 6 is a block diagram illustrating an electronic device 800 in accordance with an example embodiment. For example, the electronic device 800 may be provided as a server. Referring to fig. 6, the electronic device 800 includes a processor 822, which may be one or more in number, and a memory 832 for storing computer programs executable by the processor 822. The computer programs stored in memory 832 may include one or more modules that each correspond to a set of instructions. Further, the processor 822 may be configured to execute the computer program to perform the above-described method of acquiring a blink parameter model or fatigue detection method.
Additionally, the electronic device 800 may also include a power component 826 and a communication component 850, the power component 826 may be configured to perform power management of the electronic device 800, and the communication component 850 may be configured to enable communication, e.g., wired or wireless communication, of the electronic device 800. The electronic device 800 may also include input/output (I/O) interfaces 858. The electronic device 800 may operate based on an operating system stored in the memory 832, such as Windows Server, Mac OS XTM, UnixTM, Linux, and the like.
In another exemplary embodiment, a computer readable storage medium, such as the memory 832, is provided that includes program instructions executable by the processor 822 of the electronic device 800 to perform the above-described method of acquiring a blink parameter model or fatigue detection method.
The preferred embodiments of the present disclosure are described in detail with reference to the accompanying drawings, however, the present disclosure is not limited to the specific details of the above embodiments, and various simple modifications may be made to the technical solution of the present disclosure within the technical idea of the present disclosure, and these simple modifications all belong to the protection scope of the present disclosure. It should be noted that the various features described in the above embodiments may be combined in any suitable manner without departing from the scope of the invention. In order to avoid unnecessary repetition, various possible combinations will not be separately described in this disclosure.
In addition, any combination of various embodiments of the present disclosure may be made, and the same should be considered as the disclosure of the present disclosure, as long as it does not depart from the spirit of the present disclosure.
Claims (8)
1. A method of fatigue detection, the method comprising:
calculating a blink parameter model in the current unit time according to the obtained sampling point data in the current unit time, wherein the blink parameter model in the unit time at least comprises blink time;
when the blink parameter model in the current unit time does not meet a first preset condition, reserving the calculated blink parameter model in the current unit time; when the blink parameter model in the current unit time meets the first preset condition, ignoring the computed blink parameter model in the current unit time, and re-determining the blink parameter model in the current unit time, wherein the first preset condition is as follows: calculating that the deviation between the blink parameter model in the current unit time and an ideal blink parameter model in the unit time exceeds an acceptable level, wherein any parameter in the blink parameter model in the current unit time exceeds a parameter range formed by the parameter in the blink parameter models in the N target unit times;
determining whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time;
wherein the determining the blink parameter model for the current unit time comprises:
calculating correlation coefficients of sampling point data in each target unit time in N target unit times and sampling point data in the current unit time one by one, wherein the N target unit times are N continuous unit times before the current unit time, and N is a positive integer greater than or equal to 2;
inputting M groups of sampling point data with the maximum correlation with the sampling point data in the current unit time into an input layer of a blink parameter prediction model to predict the blink parameter model, wherein the input layer of the blink parameter prediction model comprises M input neurons, the M groups of sampling point data correspond to the M input neurons one by one, and for each input neuron, a correlation coefficient between the sampling point data corresponding to the input neuron and the sampling point data in the current unit time is used as a weight between the input neuron and a target intermediate neuron, wherein the target intermediate neuron is any neuron in a first-level hidden layer of the blink parameter prediction model, and M is more than or equal to 2 and less than or equal to N;
and finally determining the blink parameter model output from the output layer of the blink parameter prediction model as the blink parameter model in the current unit time.
2. The method of claim 1, wherein the N target unit times comprise a unit time previous to the current unit time.
3. The method of claim 1, wherein determining whether the detected user is in fatigue based on at least the blink parameter model for the current unit of time comprises:
taking the next unit time of the current unit time as a new current unit time, returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, and determining that the detected user is in a fatigue state, wherein the second preset condition is as follows: when P continuous unit time periods exist, the blink time is larger than or equal to the blink time threshold, and P is a positive integer larger than or equal to 2.
4. The method of claim 1, wherein determining whether the detected user is in fatigue based on at least the blink parameter model for the current unit of time comprises:
taking the next unit time of the current unit time as a new current unit time, and returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time until a second preset condition is met, wherein the second preset condition is as follows: when the blink time in the blink parameter model in P continuous unit time is larger than or equal to the blink time threshold value, P is a positive integer larger than or equal to 2;
determining a blink parameter model for each unit time by using the method of claim 1 or 2 one by one from the next unit time of the P unit times;
and when no blink parameter model with the blink time smaller than the blink time threshold value exists in Q unit times from the next unit time of the P unit times, determining that the detected user is in the fatigue state, wherein Q is a positive integer larger than or equal to 2.
5. The method of claim 4, wherein determining whether the detected user is in fatigue based on at least the blink parameter model for the current unit of time further comprises:
when a blink parameter model with blink time smaller than the blink time threshold exists in the Q unit time, determining that the detected user is not in a fatigue state;
the method further comprises the following steps:
and taking the next unit time of the unit time corresponding to the blink parameter model with the blink time smaller than the blink time threshold value in the Q unit times as new current unit time, and returning to the step of calculating the blink parameter model in the current unit time according to the obtained sampling point data in the current unit time.
6. A fatigue detection apparatus, characterized in that the apparatus comprises:
the second calculation module is used for calculating a blink parameter model in the current unit time according to the obtained sampling point data in the current unit time, wherein the blink parameter model in the unit time at least comprises blink time;
the second determining module is used for reserving the calculated blink parameter model in the current unit time when the blink parameter model in the current unit time does not meet a first preset condition; when the blink parameter model in the current unit time meets the first preset condition, ignoring the blink parameter model in the current unit time calculated by the second calculation module, and re-determining the blink parameter model in the current unit time by using a device for acquiring the blink parameter model, wherein the first preset condition is as follows: the second calculation module calculates that the deviation between the blink parameter model in the current unit time and the ideal blink parameter model in the unit time exceeds an acceptable level, and any parameter in the blink parameter model in the current unit time exceeds a parameter range formed by the parameter in the blink parameter models in the N target unit times;
a third determining module, configured to determine whether the detected user is in a fatigue state at least according to the blink parameter model in the current unit time;
wherein the device for acquiring the blink parameter model comprises:
the first calculation module is used for calculating correlation coefficients of sampling point data in each target unit time in N target unit times and sampling point data in the current unit time one by one, wherein the N target unit times are N continuous unit times before the current unit time, and N is a positive integer greater than or equal to 2;
the prediction module is used for inputting M groups of sampling point data with the maximum correlation with the sampling point data in the current unit time into an input layer of a blink parameter prediction model to predict the blink parameter model, wherein the input layer of the blink parameter prediction model comprises M input neurons, the M groups of sampling point data correspond to the M input neurons one by one, and for each input neuron, a correlation coefficient between the sampling point data corresponding to the input neuron and the sampling point data in the current unit time is used as a weight between the input neuron and a target intermediate neuron, wherein the target intermediate neuron is any neuron in a first-level hidden layer of the blink parameter prediction model, and M is more than or equal to 2 and less than or equal to N;
a first determining module, configured to finally determine the blink parameter model output from the output layer of the blink parameter prediction model as the blink parameter model in the current unit time.
7. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method of claims 1 to 5.
8. An electronic device, comprising:
the computer-readable storage medium recited in claim 7; and
one or more processors to execute the program in the computer-readable storage medium.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711466152.0A CN108245171B (en) | 2017-12-28 | 2017-12-28 | Method for obtaining parameter model, fatigue detection method and device, medium and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711466152.0A CN108245171B (en) | 2017-12-28 | 2017-12-28 | Method for obtaining parameter model, fatigue detection method and device, medium and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108245171A CN108245171A (en) | 2018-07-06 |
| CN108245171B true CN108245171B (en) | 2020-12-25 |
Family
ID=62724436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711466152.0A Active CN108245171B (en) | 2017-12-28 | 2017-12-28 | Method for obtaining parameter model, fatigue detection method and device, medium and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108245171B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109259724B (en) * | 2018-08-30 | 2021-11-12 | Oppo广东移动通信有限公司 | Eye monitoring method and device, storage medium and wearable device |
| CN109616201B (en) * | 2018-11-09 | 2023-10-31 | 深圳职业技术大学 | Fatigue early warning method based on small data dynamic prediction model |
| CN112702194B (en) * | 2020-12-16 | 2023-04-07 | 中国联合网络通信集团有限公司 | Indoor cell fault positioning method and device and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09218375A (en) * | 1996-02-08 | 1997-08-19 | Canon Inc | Fatigue deciding method and observing device using same |
| WO2009128562A1 (en) * | 2008-04-16 | 2009-10-22 | スカラ株式会社 | Fatigue prevention device |
| CN101583313A (en) * | 2007-01-19 | 2009-11-18 | 旭化成株式会社 | Awake state judging model making device, awake state judging device, and warning device |
| CN106919913A (en) * | 2017-02-21 | 2017-07-04 | 上海蔚来汽车有限公司 | Method for detecting fatigue driving and device based on computer vision |
| CN107440714A (en) * | 2017-09-07 | 2017-12-08 | 哈尔滨理工大学 | A kind of focus and blink action extraction element and method |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102008007152B4 (en) * | 2007-03-30 | 2019-01-17 | Volkswagen Ag | Method for the parameterization of the eye opening degree of a driver of a motor vehicle |
| CN101642374A (en) * | 2008-08-04 | 2010-02-10 | 南京大学 | Device and method for detecting fatigue |
| CN104598993B (en) * | 2015-01-14 | 2017-11-03 | 南京信息工程大学 | Driver tired driving Forecasting Methodology based on number of winks |
| US10016130B2 (en) * | 2015-09-04 | 2018-07-10 | University Of Massachusetts | Eye tracker system and methods for detecting eye parameters |
| CN106384096B (en) * | 2016-09-20 | 2018-07-10 | 西安科技大学 | A kind of fatigue driving monitoring method based on blink detection |
-
2017
- 2017-12-28 CN CN201711466152.0A patent/CN108245171B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09218375A (en) * | 1996-02-08 | 1997-08-19 | Canon Inc | Fatigue deciding method and observing device using same |
| CN101583313A (en) * | 2007-01-19 | 2009-11-18 | 旭化成株式会社 | Awake state judging model making device, awake state judging device, and warning device |
| WO2009128562A1 (en) * | 2008-04-16 | 2009-10-22 | スカラ株式会社 | Fatigue prevention device |
| CN106919913A (en) * | 2017-02-21 | 2017-07-04 | 上海蔚来汽车有限公司 | Method for detecting fatigue driving and device based on computer vision |
| CN107440714A (en) * | 2017-09-07 | 2017-12-08 | 哈尔滨理工大学 | A kind of focus and blink action extraction element and method |
Non-Patent Citations (2)
| Title |
|---|
| Eye-blink rate detection for fatigue determination;Z. A. Haq and Z. Hasan;《2016 1st India International Conference on Information Processing (IICIP)》;20170713;第1-5页 * |
| 驾驶员疲劳检测实时控制系统设计;唐新星等;《制造业自动化》;20161125;第44-51页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108245171A (en) | 2018-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108245171B (en) | Method for obtaining parameter model, fatigue detection method and device, medium and equipment | |
| CN111656362B (en) | Cognitive and occasional depth plasticity based on acoustic feedback | |
| JP2020047285A (en) | Training neural network using prioritized experience memory | |
| CN109697207A (en) | The abnormality monitoring method and system of time series data | |
| US20160239592A1 (en) | Data-driven battery aging model using statistical analysis and artificial intelligence | |
| CN110890998B (en) | Method and device for determining threshold | |
| CN111541626A (en) | Network bandwidth updating method and device, electronic equipment and storage medium | |
| KR20170015097A (en) | Method and apparatus for power management | |
| CN109800858B (en) | Application system abnormality detection method and device, readable storage medium and electronic equipment | |
| CN112163280B (en) | Method, device and equipment for simulating automatic driving scene and storage medium | |
| CN115755564A (en) | Alarm clock control method based on sleep stage prediction, radar and storage medium | |
| EP3144705A1 (en) | Information processing device, anomaly detection method and recording medium | |
| JP6277638B2 (en) | Signal processing apparatus, signal processing method, and computer program | |
| CN115390161A (en) | Precipitation prediction method and device based on artificial intelligence | |
| CN114091611A (en) | Equipment load weight obtaining method and device, storage medium and electronic equipment | |
| CN113079033B (en) | Flow control method and device, electronic equipment and computer readable medium | |
| AU2012222079B2 (en) | Customizable policy engine | |
| CN107729144B (en) | Application control method and device, storage medium and electronic equipment | |
| CN111930602B (en) | Performance index prediction method and device | |
| JP5338705B2 (en) | Exercise state detection method and exercise state detection device | |
| CN111637975A (en) | Wrist temperature measuring method and device, wearable device and storage medium | |
| US11614560B2 (en) | Integration of physical sensors in a data assimilation framework | |
| CN108171118B (en) | Blink signal data processing method and device, readable storage medium and electronic equipment | |
| CN111258863A (en) | Data anomaly detection method, device, server and computer-readable storage medium | |
| US20180357528A1 (en) | Control system, system and computer-readable medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |