CN113679396A - Training method, device, terminal and medium for fatigue recognition model - Google Patents
Training method, device, terminal and medium for fatigue recognition model Download PDFInfo
- Publication number
- CN113679396A CN113679396A CN202110930200.7A CN202110930200A CN113679396A CN 113679396 A CN113679396 A CN 113679396A CN 202110930200 A CN202110930200 A CN 202110930200A CN 113679396 A CN113679396 A CN 113679396A
- Authority
- CN
- China
- Prior art keywords
- users
- fatigue
- electroencephalogram signals
- determining
- electroencephalogram
- 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.)
- Pending
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
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
- A61B5/372—Analysis of electroencephalograms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
- A61B5/372—Analysis of electroencephalograms
- A61B5/374—Detecting the frequency distribution of signals, e.g. detecting delta, theta, alpha, beta or gamma waves
-
- 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
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Surgery (AREA)
- Public Health (AREA)
- Pathology (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Psychiatry (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Veterinary Medicine (AREA)
- Psychology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Signal Processing (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The application discloses a training method, a training device, a training terminal and a training medium for a fatigue recognition model. The method comprises the following steps: acquiring electroencephalogram signals corresponding to a plurality of first users to be trained; determining power spectrums of the electroencephalogram signals respectively corresponding to the first users according to the electroencephalogram signals respectively corresponding to the first users; determining fatigue parameters of the electroencephalogram signals respectively corresponding to the first users based on the power spectrums of the electroencephalogram signals respectively corresponding to the first users; determining fatigue grades of electroencephalogram signals respectively corresponding to a plurality of first users; and training a preset neural network model according to fatigue grades and fatigue parameters corresponding to the electroencephalogram signals respectively corresponding to the first users to obtain a fatigue recognition model. The method and the device achieve the purpose of detecting the fatigue grade of the detected user through the electroencephalogram signal, further expand the application scene of the fatigue identification model, and improve the accuracy of fatigue detection.
Description
Technical Field
The application relates to the technical field of computers, in particular to a training method, a device, a terminal and a medium for a fatigue recognition model.
Background
Fatigue refers to the subjective uncomfortable feeling of fatigue, but objectively loses the ability to complete the original normal activities or work under the same conditions. The operator works in a fatigue state, so that accidents are easily caused, particularly in the fields of driving, construction site construction and the like. In the related art, only whether the user to be tested is tired can be detected, but the specific fatigue level of the user to be tested cannot be given. Therefore, a scheme for automatically identifying the fatigue state of the tested user and the corresponding fatigue level is needed to be invented.
Disclosure of Invention
In order to solve at least one technical problem, the present application provides a training method, an apparatus, a terminal and a medium for a fatigue recognition model.
According to a first aspect of the present application, there is provided a training method of a fatigue recognition model, the method comprising:
acquiring electroencephalogram signals corresponding to a plurality of first users to be trained;
determining power spectrums of the electroencephalogram signals respectively corresponding to the first users according to the electroencephalogram signals respectively corresponding to the first users;
determining fatigue parameters of the electroencephalogram signals respectively corresponding to the first users based on the power spectrums of the electroencephalogram signals respectively corresponding to the first users;
determining fatigue grades of electroencephalogram signals respectively corresponding to a plurality of first users;
and training a preset neural network model according to the fatigue grades and the fatigue parameters of the electroencephalogram signals respectively corresponding to the first users to obtain a fatigue recognition model.
According to a second aspect of the present application, there is provided a training apparatus for a fatigue recognition model, the apparatus comprising:
the electroencephalogram signal acquisition module is used for acquiring electroencephalogram signals corresponding to a plurality of first users to be trained respectively;
the frequency domain characteristic extraction module is used for determining power spectrums of the electroencephalograms corresponding to the first users according to the electroencephalograms corresponding to the first users respectively;
the fatigue parameter determining module is used for determining fatigue parameters of the electroencephalogram signals corresponding to the first users respectively based on the power spectrums of the electroencephalogram signals corresponding to the first users respectively;
the fatigue grade determining module is used for determining the fatigue grades of the electroencephalogram signals respectively corresponding to the first users;
and the recognition model training module is used for training a preset neural network model according to the fatigue grades and the fatigue parameters of the electroencephalogram signals respectively corresponding to the first users to obtain a fatigue recognition model.
According to a third aspect of the present application, there is provided a terminal comprising: the fatigue recognition model training device comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the computer program to realize the fatigue recognition model training method.
According to a fourth aspect of the present application, there is provided a computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, is adapted to carry out the above-mentioned training method of the fatigue recognition model.
The embodiment of the application obtains the electroencephalograms corresponding to a plurality of first users to be trained respectively, determines the power spectrums corresponding to the plurality of first electroencephalograms respectively according to the electroencephalograms corresponding to the plurality of first users respectively, determines the fatigue parameters of the electroencephalograms corresponding to the plurality of first users respectively based on the power spectrums of the electroencephalograms corresponding to the plurality of first electroencephalograms respectively, determines the fatigue grades of the electroencephalograms corresponding to the plurality of first users respectively, trains the preset neural network model according to the fatigue grades and the fatigue parameters of the electroencephalograms corresponding to the plurality of first users respectively, obtains the fatigue recognition model, and trains the neural network model by determining the fatigue parameters and the fatigue grades of the electroencephalograms corresponding to different users respectively, so that the fatigue recognition model has the function of automatically recognizing the power spectrums, thereby further achieving the purpose of detecting the fatigue grades of the detected users through the electroencephalograms, further, the application scene of the fatigue identification model is expanded, and the accuracy of fatigue detection is improved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings used in the description of the embodiments of the present application will be briefly described below.
Fig. 1 is a schematic flowchart of a training method for a fatigue recognition model according to an embodiment of the present disclosure;
FIG. 2 is a schematic flowchart of another training method for a fatigue recognition model according to an embodiment of the present disclosure; and
fig. 3 is a block diagram structural schematic diagram of a training apparatus for a fatigue recognition model according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
It should be noted that although functional blocks are partitioned in a schematic diagram of an apparatus and a logical order is shown in a flowchart, in some cases, the steps shown or described may be performed in a different order than the partitioning of blocks in the apparatus or the order in the flowchart.
To make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.
According to an embodiment of the present application, there is provided a training method of a fatigue recognition model, as shown in fig. 1, the method includes steps S101 to S105.
Step S101: acquiring electroencephalogram signals corresponding to a plurality of first users to be trained respectively.
Specifically, the electronic device acquires a plurality of first electroencephalogram signals to be trained. The electronic device may be a brain electrical signal acquisition device (such as an EEG acquisition device), a brain computer interface BCI device, a mobile phone, a tablet, a PC, a server, or the like. For example, if the electronic device is a brain-computer interface BCI device, the brain-computer interface BCI device may be connected to the EEG acquisition device, so as to acquire the electroencephalogram signal acquired by the EEG acquisition device.
Specifically, the electronic device may acquire a plurality of first electroencephalogram signals to be trained according to a preset sampling frequency.
In the embodiment of the application, the electroencephalogram signals respectively corresponding to a plurality of first users are used for representing the electroencephalogram signals of the users without fatigue degrees. Wherein, the fatigue degree can be divided according to the business requirement.
Step S102: and determining power spectrums corresponding to the first electroencephalograms according to the electroencephalograms corresponding to the first users.
Specifically, multiple frequency domain feature extraction algorithms locally arranged in the electronic device unit may be preconfigured to determine the power spectrums corresponding to the first electroencephalogram signals, respectively. When the method is applied, one frequency domain feature extraction algorithm can be set as a default algorithm, and feature extraction processing on electroencephalogram signals respectively corresponding to a plurality of first users can be determined according to detected selected operations aiming at various frequency domain feature extraction algorithms.
Specifically, the electroencephalogram signals corresponding to the first users can be sent to a predetermined algorithm server, so that the purpose of extracting frequency domain features of the electroencephalogram signals corresponding to the first users by using the algorithm server is achieved.
Step S103: and determining fatigue parameters of the electroencephalogram signals respectively corresponding to the first users based on the power spectrums of the electroencephalogram signals respectively corresponding to the first electroencephalogram signals.
In the embodiment of the application, the fatigue parameter is used for representing a certain waveform or a ratio of certain waveforms to the whole waveform in the electroencephalogram signal.
Specifically, the power spectrums of the electroencephalogram signals corresponding to the first electroencephalogram signals are processed according to a preset fatigue parameter algorithm, so that fatigue parameters of the electroencephalogram signals corresponding to the first electroencephalogram signals are obtained.
Step S104: and determining the fatigue grades of the electroencephalogram signals respectively corresponding to the first users.
In the present embodiment, the fatigue level is used to characterize the degree of fatigue. When the method is applied, the division can be carried out according to business requirements (such as the requirement on fatigue identification precision). For example, the fatigue level may include deep fatigue, shallow fatigue, wakefulness, and the like.
Specifically, the fatigue level is generally determined according to the labeling of the electroencephalogram signal.
Step S105: and training a preset neural network model according to the fatigue grades and the fatigue parameters of the electroencephalogram signals respectively corresponding to the first users to obtain a fatigue recognition model.
Specifically, the preset neural network model may be a random forest model, an SVM support vector machine, a DNN deep neural network, or the like.
Specifically, the fatigue grade and the fatigue parameter of the electroencephalogram signal corresponding to each of the first users can be divided into two parts, one part is training data, the other part is verification data, and each parameter of the neural network model is adjusted through training of the two parts of data, so that the neural network model after the parameters are adjusted has high identification precision.
The embodiment of the application obtains the electroencephalograms corresponding to a plurality of first users to be trained respectively, determines the power spectrums corresponding to the plurality of first electroencephalograms respectively according to the electroencephalograms corresponding to the plurality of first users respectively, determines the fatigue parameters of the electroencephalograms corresponding to the plurality of first users respectively based on the power spectrums of the electroencephalograms corresponding to the plurality of first electroencephalograms respectively, determines the fatigue grades of the electroencephalograms corresponding to the plurality of first users respectively, trains the preset neural network model according to the fatigue grades and the fatigue parameters of the electroencephalograms corresponding to the plurality of first users respectively, obtains the fatigue recognition model, and trains the neural network model by determining the fatigue parameters and the fatigue grades of the electroencephalograms corresponding to different users respectively, so that the fatigue recognition model has the function of automatically recognizing the power spectrums, thereby further achieving the purpose of detecting the fatigue grades of the detected users through the electroencephalograms, further, the application scene of the fatigue identification model is expanded, and the accuracy of fatigue detection is improved.
In some embodiments, step S104 includes at least one of the following sub-steps:
determining fatigue levels of the electroencephalograms corresponding to the first users respectively based on marking operations of the electroencephalograms corresponding to the first users respectively;
and determining fatigue grades of the electroencephalograms corresponding to the first users respectively based on the pre-labeled fatigue grade identifications of the electroencephalograms corresponding to the first users respectively.
Specifically, the marking processing of different electroencephalogram signals can be provided for a user through a preset interactive interface, and the marking mode through the interactive interface plays a role in marking the electroencephalogram signals in real time, so that the problem that the preset neural network model cannot be trained due to unmarked or label missing of the electroencephalogram signals is solved. During application, when the marking operation aiming at the electroencephalogram signals respectively corresponding to a plurality of first users and input by the user is detected, the fatigue grades of the electroencephalogram signals respectively corresponding to the plurality of first users are determined.
According to the embodiment of the application, through the pre-marking processing, the electroencephalogram signals acquired by the electronic equipment in the step S101 are provided with the mark records, and therefore the steps of determining the fatigue level subsequently are reduced.
In some embodiments, step S102 includes:
and respectively converting the electroencephalogram signals respectively corresponding to the first users into corresponding power spectrums based on a preset local frequency domain feature extraction algorithm.
Specifically, the frequency domain feature extraction algorithm is preset locally to the electronic device.
Specifically, the frequency domain feature extraction algorithm preset locally may be a fourier transform method.
In some embodiments, before step S102, the method further comprises:
and filtering, denoising, removing pseudo and normalizing the electroencephalogram signals corresponding to the first users respectively.
Specifically, a notch filtering algorithm can be used for removing 50Hz power frequency interference, and then a preset filter is used for filtering high-frequency signals and low-frequency signals in the electroencephalogram signals. For example, if the frequency setting range of the filter is: 0.5-35 HZ, then the high frequency part which is larger than 35HZ and the low frequency part which is smaller than 5HZ can be filtered by the filter.
Specifically, a recursive average noise algorithm, a minimum tracking algorithm, a histogram noise estimation algorithm, and the like may be used to perform the drying processing on the electroencephalogram signals.
Specifically, a principal component analysis method can be used for removing interference waveforms generated by blink myoelectricity with certain pattern characteristics, and therefore, the electroencephalogram signals can be subjected to de-counterfeiting processing.
In particular, the normalization may be done with the volt-value averaged over every 10s, i.e. normalized per 10s by the mean EEG intensity.
In some embodiments, before step S102, the method further comprises:
the electroencephalogram signals respectively corresponding to the first users are subjected to segmentation processing to obtain a plurality of segment electroencephalogram signals respectively corresponding to the first users, and power spectrums respectively corresponding to the plurality of segment electroencephalogram signals respectively corresponding to the first users are respectively determined.
Specifically, the electroencephalogram signal may be subjected to the segmentation processing at predetermined time intervals. For example, assuming that the time duration of the electroencephalogram signal corresponding to each first user is 10 seconds, the segmentation processing may be performed according to a predetermined time duration segment (e.g., 2 seconds), so that each segment of electroencephalogram signal is processed according to steps S102-S105.
In some embodiments, as shown in fig. 1, step S103 further comprises:
step S1031: determining a theta wave power spectrum, an alpha wave power spectrum and a beta wave power spectrum of the electroencephalogram signals respectively corresponding to the first users based on the power spectrums of the electroencephalogram signals respectively corresponding to the first users;
step S1032: determining fatigue parameters of the electroencephalogram signals respectively corresponding to the first users according to the theta wave power spectrum, the alpha wave power spectrum and the beta wave power spectrum of the electroencephalogram signals respectively corresponding to the first users;
the fatigue parameter of the electroencephalogram signal corresponding to any first user comprises the proportion of the sum of the theta wave power spectrum and the alpha wave power spectrum of the electroencephalogram signal corresponding to any first user in the power spectrum of the electroencephalogram signal corresponding to any first user, and the proportion of the beta wave power spectrum of the electroencephalogram signal corresponding to any first user in the power spectrum of the electroencephalogram signal corresponding to any first user.
In some embodiments, as shown in fig. 2, the method further comprises:
step S201: acquiring an electroencephalogram signal of a second user to be identified;
step S202: converting the electroencephalogram signal of the second user into a target power spectrum;
step S203: determining a fatigue parameter for the brain electrical signal of the second user based on the target power spectrum;
step S204: and determining the fatigue grade corresponding to the fatigue parameter of the electroencephalogram signal of the second user according to the fatigue identification model.
Specifically, the second brain electrical signal may be converted into a target power spectrum with reference to step S102.
Specifically, before the electroencephalogram signal of the second user is converted into the target power spectrum, filtering, noise reduction, artifact removal and normalization processing can be further performed on the electroencephalogram signal of the second user.
Specifically, the fatigue parameter of the brain electrical signal for the second user may be determined with reference to steps S1031 and S1032.
According to the fatigue detection method and device, the fatigue identification model is used, the effect of detecting the fatigue level of the electroencephalogram signal is achieved, the time for detecting the fatigue level is shortened, and the fatigue detection efficiency is improved.
In another embodiment of the present application, there is provided a training apparatus for a fatigue recognition model, as shown in fig. 3, the apparatus 30 includes: the system comprises an electroencephalogram signal acquisition module 301, a frequency domain feature extraction module 302, a fatigue parameter determination module 303, a fatigue grade determination module 304 and a recognition model training module 305.
The electroencephalogram signal acquisition module 301 is configured to acquire electroencephalogram signals corresponding to a plurality of first users to be trained;
the frequency domain feature extraction module 302 is configured to extract frequency domain features of the electroencephalogram signals corresponding to the multiple first users, so as to obtain power spectrums of the electroencephalogram signals corresponding to the multiple first users;
the fatigue parameter determining module 303 is configured to determine fatigue parameters of the electroencephalogram signals corresponding to the plurality of first users based on power spectrums of the electroencephalogram signals corresponding to the plurality of first users, respectively;
the fatigue level determining module 304 is configured to determine fatigue levels to which electroencephalogram signals corresponding to a plurality of first users belong;
the recognition model training module 305 is configured to train a preset neural network model according to fatigue grades and fatigue parameters to which electroencephalogram signals respectively corresponding to a plurality of first users belong, so as to obtain a fatigue recognition model.
The embodiment of the application obtains the electroencephalograms corresponding to a plurality of first users to be trained respectively, determines the power spectrums corresponding to the plurality of first electroencephalograms respectively according to the electroencephalograms corresponding to the plurality of first users respectively, determines the fatigue parameters of the electroencephalograms corresponding to the plurality of first users respectively based on the power spectrums of the electroencephalograms corresponding to the plurality of first electroencephalograms respectively, determines the fatigue grades of the electroencephalograms corresponding to the plurality of first users respectively, trains the preset neural network model according to the fatigue grades and the fatigue parameters of the electroencephalograms corresponding to the plurality of first users respectively, obtains the fatigue recognition model, and trains the neural network model by determining the fatigue parameters and the fatigue grades of the electroencephalograms corresponding to different users respectively, so that the fatigue recognition model has the function of automatically recognizing the power spectrums, thereby further achieving the purpose of detecting the fatigue grades of the detected users through the electroencephalograms, further, the application scene of the fatigue identification model is expanded, and the accuracy of fatigue detection is improved.
Further, the fatigue level determination module comprises at least one of the following sub-modules:
the first determining submodule is used for determining fatigue grades of the electroencephalograms corresponding to the first users based on marking operation of the electroencephalograms corresponding to the first users;
and the second determining submodule is used for determining the fatigue levels of the electroencephalograms corresponding to the first users respectively based on the fatigue level identifications of the electroencephalograms corresponding to the pre-labeled first users respectively.
Further, the frequency domain feature extraction module comprises:
and the characteristic extraction submodule is used for converting the extracted electroencephalogram signals respectively corresponding to the first users into corresponding power spectrums respectively based on a preset local frequency domain characteristic extraction algorithm.
Further, before determining the power spectrums of the electroencephalograms corresponding to the plurality of first users according to the electroencephalograms corresponding to the plurality of first users, the frequency domain feature extraction module further includes:
and the signal preprocessing submodule is used for filtering, denoising, pseudo-removing and normalizing the extracted electroencephalogram signals respectively corresponding to the first users.
Further, before the step of determining the power spectrums of the electroencephalograms corresponding to the plurality of first users according to the electroencephalograms corresponding to the plurality of first users, the frequency domain feature extraction module further includes:
the segmentation processing submodule is used for performing segmentation processing on the extracted electroencephalograms corresponding to the first users respectively to obtain a plurality of segment electroencephalograms corresponding to the first users respectively, so that power spectrums corresponding to the plurality of segment electroencephalograms corresponding to the first users respectively are determined respectively.
Further, the fatigue parameter determination module comprises:
the designated wave spectrum determination submodule is used for determining a theta wave power spectrum, an alpha wave power spectrum and a beta wave power spectrum of the electroencephalogram signals respectively corresponding to the first users based on the power spectrums of the electroencephalogram signals respectively corresponding to the first users;
the fatigue parameter calculation submodule is used for determining fatigue parameters of the electroencephalogram signals of the first users according to the theta wave power spectrum, the alpha wave power spectrum and the beta wave power spectrum of the electroencephalogram signals respectively corresponding to the first users;
the fatigue parameters of the electroencephalogram signal of any first user comprise the proportion of the sum of the theta wave power spectrum and the alpha wave power spectrum of the electroencephalogram signal of any first user in the power spectrum of the electroencephalogram signal of any first user, and the proportion of the beta wave power spectrum of the electroencephalogram signal of any first user in the power spectrum of the electroencephalogram signal of any first user.
Further, the apparatus further comprises:
the signal to be identified acquisition module is used for acquiring an electroencephalogram signal of a second user to be identified;
the signal to be identified conversion module is used for converting the electroencephalogram signal of the second user into a target power spectrum;
the to-be-processed fatigue parameter determining module is used for determining a fatigue parameter of the electroencephalogram signal of the second user based on the target power spectrum;
and the fatigue grade determining module is used for determining the fatigue grade of the fatigue parameter aiming at the electroencephalogram signal of the second user according to the fatigue identification model.
The training device for the fatigue recognition model of the embodiment can execute the training method for the fatigue recognition model provided by the embodiment of the application, and the implementation principles are similar, and are not repeated here.
Another embodiment of the present application provides a terminal, including: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the computer program implementing the training method of the fatigue recognition model described above.
In particular, the processor may be a CPU, general purpose processor, DSP, ASIC, FPGA or other programmable logic device, transistor logic device, hardware component, or any combination thereof. Which may implement or perform the various illustrative logical blocks, modules, and circuits described in connection with the disclosure. A processor may also be a combination of computing functions, e.g., comprising one or more microprocessors, a DSP and a microprocessor, or the like.
In particular, the processor is coupled to the memory via a bus, which may include a path for communicating information. The bus may be a PCI bus or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc.
The memory may be, but is not limited to, a ROM or other type of static storage device that can store static information and instructions, a RAM or other type of dynamic storage device that can store information and instructions, an EEPROM, a CD-ROM or other optical disk storage, optical disk storage (including compact disk, laser disk, optical disk, digital versatile disk, blu-ray disk, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.
Optionally, the memory is used for storing codes of computer programs for executing the scheme of the application, and the processor is used for controlling the execution. The processor is used for executing the application program codes stored in the memory to realize the actions of the training device of the fatigue recognition model provided by the above embodiment.
Yet another embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the above-mentioned training method for a fatigue recognition model.
The above-described embodiments of the apparatus are merely illustrative, and the units illustrated as separate components may or may not be physically separate, may be located in one place, or may be distributed over a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
One of ordinary skill in the art will appreciate that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as known to those skilled in the art.
While the present invention has been described with reference to the preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A training method of a fatigue recognition model is characterized by comprising the following steps:
acquiring electroencephalogram signals corresponding to a plurality of first users to be trained;
determining power spectrums of the electroencephalogram signals respectively corresponding to the first users according to the electroencephalogram signals respectively corresponding to the first users;
determining fatigue parameters of the electroencephalogram signals respectively corresponding to the first users based on the power spectrums of the electroencephalogram signals respectively corresponding to the first users;
determining fatigue grades of electroencephalogram signals respectively corresponding to a plurality of first users;
and training a preset neural network model according to fatigue grades and fatigue parameters corresponding to the electroencephalogram signals respectively corresponding to the first users to obtain a fatigue recognition model.
2. The method of claim 1, wherein the determining the fatigue levels of the electroencephalogram signals corresponding to the plurality of first users respectively comprises at least one of the following steps:
determining fatigue levels of the electroencephalograms corresponding to the first users based on marking operations of the electroencephalograms corresponding to the first users respectively;
and determining fatigue grades of the electroencephalograms corresponding to the first users respectively based on the pre-labeled fatigue grade identifications of the electroencephalograms corresponding to the first users respectively.
3. The method of claim 1, wherein said step of determining the power spectrum of the electroencephalogram signal corresponding to each of the plurality of first users from the electroencephalogram signal corresponding to each of the plurality of first users comprises:
and respectively converting the extracted electroencephalogram signals respectively corresponding to the plurality of first users into corresponding power spectrums based on a preset local frequency domain feature extraction algorithm.
4. The method of claim 1, wherein prior to the step of determining the power spectrum of the brain electrical signal corresponding to each of the first plurality of users from the brain electrical signal corresponding to each of the first plurality of users, the method further comprises:
and filtering, denoising, pseudo-removing and normalizing the extracted electroencephalograms corresponding to the plurality of first users respectively.
5. The method of claim 1, wherein before the step of extracting the frequency domain features of the electroencephalogram signals corresponding to a plurality of the first users, the method further comprises:
and carrying out segmentation processing on the extracted electroencephalograms corresponding to the first users respectively to obtain a plurality of segment electroencephalograms corresponding to the first users respectively so as to determine respective power spectrums corresponding to the plurality of segment electroencephalograms corresponding to the first users respectively.
6. The method of claim 1, wherein the step of determining fatigue parameters for extracting the electroencephalogram signals corresponding to the plurality of first users based on extracting power spectra of the electroencephalogram signals corresponding to the plurality of first users comprises:
determining a theta wave power spectrum, an alpha wave power spectrum and a beta wave power spectrum of the electroencephalogram signals respectively corresponding to the first users based on the power spectrums of the electroencephalogram signals respectively corresponding to the first users;
determining fatigue parameters of the electroencephalogram signals of the first users according to theta wave power spectrums, alpha wave power spectrums and beta wave power spectrums of the electroencephalogram signals respectively corresponding to the first users;
the fatigue parameters of the electroencephalogram signal of any one of the first users comprise the proportion of the sum of the theta wave power spectrum and the alpha wave power spectrum of the electroencephalogram signal of any one of the first users in the power spectrum of the electroencephalogram signal of any one of the first users, and the proportion of the beta wave power spectrum of the electroencephalogram signal of any one of the first users in the power spectrum of the electroencephalogram signal of any one of the first users.
7. The method of claim 1, further comprising:
acquiring an electroencephalogram signal of a second user to be identified;
converting the electroencephalogram signal of the second user into a target power spectrum;
determining a fatigue parameter of the electroencephalogram signal of the second user based on the target power spectrum;
and determining the fatigue grade of the fatigue parameter aiming at the electroencephalogram signal of the second user according to the fatigue identification model.
8. A training device for a fatigue recognition model, comprising:
the electroencephalogram signal acquisition module is used for acquiring electroencephalogram signals corresponding to a plurality of first users to be trained respectively;
the frequency domain feature extraction module is used for determining power spectrums of the electroencephalograms corresponding to the first users according to the electroencephalograms corresponding to the first users respectively;
the fatigue parameter determining module is used for determining fatigue parameters of the electroencephalogram signals corresponding to the first users respectively based on the power spectrums of the electroencephalogram signals corresponding to the first users respectively;
the fatigue grade determining module is used for determining the fatigue grades of the electroencephalogram signals respectively corresponding to the first users;
and the recognition model training module is used for training a preset neural network model according to fatigue grades and fatigue parameters corresponding to the electroencephalogram signals respectively corresponding to the first users to obtain a fatigue recognition model.
9. A terminal, comprising: memory, processor and computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method of any of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, is adapted to carry out the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110930200.7A CN113679396A (en) | 2021-08-13 | 2021-08-13 | Training method, device, terminal and medium for fatigue recognition model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110930200.7A CN113679396A (en) | 2021-08-13 | 2021-08-13 | Training method, device, terminal and medium for fatigue recognition model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113679396A true CN113679396A (en) | 2021-11-23 |
Family
ID=78579843
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110930200.7A Pending CN113679396A (en) | 2021-08-13 | 2021-08-13 | Training method, device, terminal and medium for fatigue recognition model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113679396A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116304563A (en) * | 2023-02-09 | 2023-06-23 | 清华大学 | Construction worker fatigue degree calculation method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109480871A (en) * | 2018-10-30 | 2019-03-19 | 北京机械设备研究所 | A kind of fatigue detection method towards RSVP brain-computer interface |
| CN110215206A (en) * | 2019-06-12 | 2019-09-10 | 中国科学院自动化研究所 | Stereoscopic display visual fatigue evaluation method, system, device based on EEG signals |
| WO2020156589A1 (en) * | 2019-02-01 | 2020-08-06 | 五邑大学 | Fatigue detection method and apparatus, and storage medium |
| CN112426162A (en) * | 2020-11-23 | 2021-03-02 | 重庆邮电大学 | Fatigue detection method based on electroencephalogram signal rhythm entropy |
-
2021
- 2021-08-13 CN CN202110930200.7A patent/CN113679396A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109480871A (en) * | 2018-10-30 | 2019-03-19 | 北京机械设备研究所 | A kind of fatigue detection method towards RSVP brain-computer interface |
| WO2020156589A1 (en) * | 2019-02-01 | 2020-08-06 | 五邑大学 | Fatigue detection method and apparatus, and storage medium |
| CN110215206A (en) * | 2019-06-12 | 2019-09-10 | 中国科学院自动化研究所 | Stereoscopic display visual fatigue evaluation method, system, device based on EEG signals |
| CN112426162A (en) * | 2020-11-23 | 2021-03-02 | 重庆邮电大学 | Fatigue detection method based on electroencephalogram signal rhythm entropy |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116304563A (en) * | 2023-02-09 | 2023-06-23 | 清华大学 | Construction worker fatigue degree calculation method and system |
| CN116304563B (en) * | 2023-02-09 | 2024-02-06 | 清华大学 | Construction worker fatigue degree calculation method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111179265B (en) | Image-based fingerprint quality evaluation method and device and electronic equipment | |
| CN111956217B (en) | Blink artifact identification method and system for real-time electroencephalogram signals | |
| CN107233103B (en) | High-speed rail dispatcher fatigue state evaluation method and system | |
| US9218540B2 (en) | Apparatus and computer readable medium for signal classification using spectrogram and templates | |
| CN111150392A (en) | Directed dynamic brain function network multi-class emotion recognition construction method and device | |
| CN113486752B (en) | Emotion recognition method and system based on electrocardiosignal | |
| CN113283331A (en) | Multi-class target identification method and system for unattended sensor system | |
| CN109116243B (en) | Power battery abnormity detection method and device | |
| CN111863014A (en) | Audio processing method and device, electronic equipment and readable storage medium | |
| CN113679396A (en) | Training method, device, terminal and medium for fatigue recognition model | |
| CN106933916B (en) | JSON character string processing method and device | |
| Boubchir et al. | Time-frequency image descriptors-based features for EEG epileptic seizure activities detection and classification | |
| CN104706366A (en) | Distraction testing method, device and system | |
| CN113679386A (en) | Method, device, terminal and medium for recognizing attention | |
| CN113208632A (en) | Attention detection method and system based on convolutional neural network | |
| CN113397563A (en) | Training method, device, terminal and medium for depression classification model | |
| KR102236437B1 (en) | Method and apparatus for mental workload assessment of multitasking user using physiological data and simulated data | |
| CN111292767B (en) | Audio event detection method and device and equipment | |
| CN114627889A (en) | Multi-sound-source sound signal processing method and device, storage medium and electronic equipment | |
| CN112183540A (en) | Label parameter extraction method and device, storage medium and electronic device | |
| WO2020232641A1 (en) | Device operation data representation method and apparatus, and storage medium | |
| CN111341304A (en) | Method, device and equipment for training speech characteristics of speaker based on GAN | |
| CN112235693B (en) | Microphone signal processing method, device, equipment and computer readable storage medium | |
| CN110827811A (en) | Voice control method and device for household electrical appliance | |
| CN109784195B (en) | Fingerprint identification method and system for enterprise fingerprint card punching |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20211123 |
|
| RJ01 | Rejection of invention patent application after publication |