Disclosure of Invention
The application provides a method and a device for updating an action pattern recognition model, which are used for at least solving the technical problem of poor adaptability of the action pattern recognition model in the related technology.
According to an aspect of the embodiments of the present application, there is provided a method for updating an action pattern recognition model, including:
acquiring a first surface electromyographic signal, wherein the first surface electromyographic signal is a surface electromyographic signal to be identified, which is acquired from a first object;
the first surface electromyographic signals are identified by using a first action pattern identification model to obtain an identification result, wherein the first action pattern identification model is obtained by training an initial action pattern identification model by using a second surface electromyographic signal, and the second surface electromyographic signal is a surface electromyographic signal sample acquired from the first object;
acquiring a target surface electromyographic signal, the identification result of which meets a target condition, from the first surface electromyographic signal;
and updating the first action mode identification model by using the target surface electromyographic signals to obtain a second action mode identification model, wherein the second action mode identification model is used for carrying out action mode identification on the first object.
Optionally, identifying the first surface electromyographic signal using the first action mode identification model, and obtaining the identification result includes:
identifying the first surface electromyographic signals by using the first action mode identification model to obtain prediction labels corresponding to the first surface electromyographic signals;
and determining the confidence of the predictive label, wherein the identification result comprises the predictive label and the confidence.
Optionally, the obtaining the target surface electromyographic signal, of which the identification result meets the target condition, from the first surface electromyographic signal includes:
acquiring a third surface electromyographic signal from the first surface electromyographic signal, wherein the prediction label corresponding to the third surface electromyographic signal is a label with wrong identification;
and determining the third surface electromyographic signal as the target surface electromyographic signal.
Optionally, the obtaining the target surface electromyographic signal, of which the identification result meets the target condition, from the first surface electromyographic signal includes:
acquiring a fourth surface electromyographic signal from the first surface electromyographic signal, wherein the predictive label corresponding to the fourth surface electromyographic signal is a label with correct identification;
acquiring a fifth surface electromyographic signal, of which the corresponding confidence is lower than a target confidence, from the fourth surface electromyographic signal;
and determining the fifth surface electromyographic signal as the target surface electromyographic signal.
Optionally, after updating the first motion pattern recognition model by using the target surface electromyographic signal to obtain a second motion pattern recognition model, the method further includes:
obtaining a sixth surface electromyographic signal, wherein the sixth surface electromyographic signal is a surface electromyographic signal sample acquired from a second object;
and updating the second action pattern recognition model by using the sixth surface electromyographic signal to obtain a third action pattern recognition model, wherein the third action pattern recognition model is used for carrying out action pattern recognition on the second object.
According to another aspect of the embodiments of the present application, there is also provided an apparatus for updating an action pattern recognition model, including:
the first acquisition module is used for acquiring a first surface electromyographic signal, wherein the first surface electromyographic signal is a surface electromyographic signal to be identified acquired from a first object;
the recognition module is used for recognizing the first surface electromyographic signals by using a first action pattern recognition model to obtain a recognition result, wherein the first action pattern recognition model is obtained by training an initial action pattern recognition model by using a second surface electromyographic signal, and the second surface electromyographic signal is a surface electromyographic signal sample acquired from the first object;
the second acquisition module is used for acquiring the target surface electromyographic signals, the identification results of which meet the target conditions, from the first surface electromyographic signals;
the first updating module is used for updating the first action pattern recognition model by using the target surface electromyographic signals to obtain a second action pattern recognition model, wherein the second action pattern recognition model is used for carrying out action pattern recognition on the first object.
Optionally, the identification module includes:
the identification unit is used for identifying the first surface electromyographic signals by using the first action mode identification model to obtain prediction labels corresponding to the first surface electromyographic signals;
and the first determining unit is used for determining the confidence coefficient of the predictive label, wherein the identification result comprises the predictive label and the confidence coefficient.
Optionally, the apparatus further comprises:
the third acquisition module is used for acquiring a sixth surface electromyographic signal after updating the first action pattern recognition model by using the target surface electromyographic signal to obtain a second action pattern recognition model, wherein the sixth surface electromyographic signal is a surface electromyographic signal sample acquired from a second object;
and the second updating module is used for updating the second action pattern recognition model by using the sixth surface electromyographic signal to obtain a third action pattern recognition model, wherein the third action pattern recognition model is used for carrying out action pattern recognition on the second object.
According to another aspect of the embodiments of the present application, there is also provided a storage medium including a stored program that when executed performs the above-described method.
According to another aspect of the embodiments of the present application, there is also provided an electronic device including a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor executing the method described above by the computer program.
In the embodiment of the application, acquiring a first surface electromyographic signal, wherein the first surface electromyographic signal is a surface electromyographic signal to be identified acquired from a first object; the method comprises the steps of using a first action pattern recognition model to recognize a first surface electromyographic signal to obtain a recognition result, wherein the first action pattern recognition model is obtained by training an initial action pattern recognition model by using a second surface electromyographic signal, and the second surface electromyographic signal is a surface electromyographic signal sample acquired from a first object; acquiring a target surface electromyographic signal with the identification result meeting a target condition from the first surface electromyographic signal; and updating the first action pattern recognition model by using the target surface electromyographic signals to obtain a second action pattern recognition model, wherein the second action pattern recognition model is used for recognizing the action pattern of the first object, and in the process of recognizing the first surface electromyographic signals to be recognized of the first object by using the first action pattern recognition model obtained through training, the target surface electromyographic signals with recognition results meeting target conditions are collected to update the first action pattern recognition model to obtain the second action pattern recognition model, so that the aim of updating the model in real time by adopting an unsupervised mode in the recognition process is fulfilled, and the adaptability of the model to the first object is improved, thereby realizing the technical effect of improving the adaptability of the action pattern recognition model, and further solving the technical problem of poor adaptability of the action pattern recognition model in the related technology.
Detailed Description
In order to make the present application solution better understood by those skilled in the art, the following description will be made in detail and with reference to the accompanying drawings in the embodiments of the present application, it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, shall fall within the scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that embodiments of the present application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to an aspect of the embodiments of the present application, a method embodiment of updating an action pattern recognition model is provided.
Alternatively, in the present embodiment, the above-described method of updating the action pattern recognition model may be applied to a hardware environment constituted by the terminal 101 and the server 103 as shown in fig. 1. As shown in fig. 1, the server 103 is connected to the terminal 101 through a network, which may be used to provide services (such as game services, application services, etc.) to the terminal or clients installed on the terminal, and a database may be provided on the server or independent of the server, for providing data storage services to the server 103, where the network includes, but is not limited to: the terminal 101 is not limited to a PC, a mobile phone, a tablet computer, or the like. The method for updating the motion pattern recognition model according to the embodiment of the present application may be performed by the server 103, may be performed by the terminal 101, or may be performed by both the server 103 and the terminal 101. The method for updating the action pattern recognition model executed by the terminal 101 according to the embodiment of the present application may be executed by a client installed thereon.
FIG. 2 is a flowchart of an alternative method of updating an action pattern recognition model according to an embodiment of the present application, as shown in FIG. 2, the method may include the steps of:
step S202, acquiring a first surface electromyographic signal, wherein the first surface electromyographic signal is a surface electromyographic signal to be identified acquired from a first object;
step S204, a first action pattern recognition model is used for recognizing the first surface electromyographic signals to obtain a recognition result, wherein the first action pattern recognition model is obtained by training an initial action pattern recognition model by using a second surface electromyographic signal, and the second surface electromyographic signal is a surface electromyographic signal sample acquired from the first object;
step S206, acquiring a target surface electromyographic signal, of which the identification result meets a target condition, from the first surface electromyographic signal;
step S208, updating the first motion pattern recognition model by using the target surface electromyographic signal to obtain a second motion pattern recognition model, where the second motion pattern recognition model is used to perform motion pattern recognition on the first object.
Through the steps S202 to S208, in the process of identifying the first surface electromyographic signals to be identified of the first object by the first action pattern identification model obtained through training, the target surface electromyographic signals with the identification result meeting the target condition are collected to update the first action pattern identification model, so as to obtain the second action pattern identification model, thereby achieving the purpose of updating the model in real time by adopting an unsupervised mode in the identification process, improving the adaptability of the model to the first object, realizing the technical effect of improving the adaptability of the action pattern identification model, and further solving the technical problem of poor adaptability of the action pattern identification model in the related art.
In the technical solution provided in step S202, the first object may include, but is not limited to, a person, an animal, an insect, and the like.
In the solution provided in step S204, the initial motion pattern recognition model may, but is not limited to, include an SVM model. The SVM is based on the idea of minimizing structural risk in a statistical learning theory, and the partition hyperplane with the maximum interval is found by solving the convex quadratic programming problem.
In the solution provided in step S206, the target condition may be, but is not limited to, any condition for finding a sample for updating the model. The target conditions may be set according to actual update requirements.
In the technical scheme provided in step S208, the updated second motion pattern recognition model is used for performing motion pattern recognition on the first object, and since the second motion pattern recognition model is integrated with more information of the first object, the recognition rate can be improved to a certain extent.
Alternatively, in the present embodiment, the action pattern recognition model can be used to recognize action patterns of respective sites, such as: gesture mode, facial mode, eye mode, leg mode, and so forth.
As an optional embodiment, the identifying the first surface electromyographic signal using the first action mode identification model, and obtaining the identification result includes:
s11, identifying the first surface electromyographic signals by using the first action mode identification model to obtain prediction labels corresponding to the first surface electromyographic signals;
s12, determining the confidence coefficient of the predictive label, wherein the identification result comprises the predictive label and the confidence coefficient.
Alternatively, in the present embodiment, the identification result may include, but is not limited to, a prediction tag obtained by identifying the first surface electromyographic signal and a confidence level corresponding to the prediction tag.
As an optional embodiment, obtaining the target surface electromyographic signal, of which the identification result meets a target condition, from the first surface electromyographic signal includes:
s21, acquiring a third surface electromyographic signal from the first surface electromyographic signal, wherein the prediction label corresponding to the third surface electromyographic signal is a label with wrong identification;
and S22, determining the third surface electromyographic signal as the target surface electromyographic signal.
Alternatively, in the present embodiment, the model may be updated with the surface electromyographic signal predicting the label error as the target surface electromyographic signal. Such as: the surface electromyographic signals with incorrect recognition are obtained to be used as candidate support vectors of the SVM model, and the current model is obviously not optimal and needs to be updated because the surface electromyographic signals cannot be predicted correctly under the current model.
As an optional embodiment, obtaining the target surface electromyographic signal, of which the identification result meets a target condition, from the first surface electromyographic signal includes:
s31, acquiring a fourth surface electromyographic signal from the first surface electromyographic signal, wherein the predicted label corresponding to the fourth surface electromyographic signal is a label with correct identification;
s32, acquiring a fifth surface electromyographic signal with the confidence lower than a target confidence from the fourth surface electromyographic signal;
and S33, determining the fifth surface electromyographic signal as the target surface electromyographic signal.
Alternatively, in the present embodiment, for predicting the correct surface electromyographic signal, the confidence thereof may be determined, and the surface electromyographic signal having the confidence lower than the target confidence may be taken as the target surface electromyographic signal. For example: if the surface electromyographic signals are correctly identified, but are actually within the two boundaries of the current classification hyperplane, the resulting model is still not optimal. Therefore, the confidence level of the surface electromyographic signals needs to be calculated, a threshold value can be obtained empirically to serve as a target confidence level, the confidence level is compared with the threshold value, and when the confidence level is smaller than the threshold value, the surface electromyographic signals are considered to be on the classification hyperplane and can serve as candidate support vectors to update the model.
In an alternative embodiment, fig. 3 is a schematic diagram of an update process of an action pattern recognition model according to an alternative embodiment of the present application, as shown in fig. 3, model learning is started for an initial training set by training set data, an initial support vector machine SVM classification model is trained, a model classifier is obtained, and then samples tested online are sent to the classifier one by one as label-free recognition data, and confidence is calculated to perform judgment while predicting and classifying: if the current sample is wrongly identified, the new support vector is considered, the sample is used for updating the current classifier model subsequently, otherwise, whether the confidence coefficient meets the threshold requirement is judged, and if the confidence coefficient meets the threshold requirement, the new support vector is considered for updating the current classifier model.
As an optional embodiment, after updating the first motion pattern recognition model with the target surface electromyographic signal to obtain a second motion pattern recognition model, the method further includes:
s41, acquiring a sixth surface electromyographic signal, wherein the sixth surface electromyographic signal is a surface electromyographic signal sample acquired from a second object;
s42, updating the second action pattern recognition model by using the sixth surface electromyographic signal to obtain a third action pattern recognition model, wherein the third action pattern recognition model is used for carrying out action pattern recognition on the second object.
Alternatively, in the present embodiment, when the subject is changed from the first subject to the second subject, model updating is performed using the sixth surface electromyographic signal of the second subject to accommodate model deviation caused by individual difference. For example: as shown in fig. 3, the SVM model may be updated with the surface electromyographic signals of the second object as labeled identification data.
Optionally, in this embodiment, the above process may be used in a scenario where gesture motion recognition of an myoelectric bracelet is used to control a bionic prosthesis to execute a corresponding motion instruction.
The application also provides an alternative embodiment, the specificity of the surface electromyographic signals is very large, and in the process of performing motion mode recognition, the performance of the motion recognition system is affected by the individual difference of the surface electromyographic signals, the individual difference is often caused by the electrode displacement of different collecting testees or the difference of the shape and the size of muscles, the strength, the fatigue degree and the skin impedance among different testees, and the individual difference often causes the training data and the test data from different testees to have different distributions, so that the classifier model obtained from the current individual learning is difficult to effectively expand and apply to other individuals.
The alternative embodiment expands the existing static classifier into a dynamic model with incremental learning capability, so that the classifier has online learning capability, can learn more and more effective information from unlabeled or marked samples, and updates the classification model in real time to enhance the self-adaptability of myoelectric action pattern recognition so as to keep the success rate of action pattern recognition of the myoelectric signals under the influence of long-term time-varying factors and individual differences from being reduced.
FIG. 4 is a schematic illustration of an update process of an action pattern recognition model, according to an alternative embodiment of the present application, as shown in FIG. 4, including the following stages:
in the training stage, firstly, recognition data set extraction is carried out, surface electromyographic signals are preprocessed and feature extraction is carried out, and feature vectors are randomly divided into two groups: a training set and a testing set; and then establishing an initial model, training an SVM classifier based on the training set data and the sample label to obtain an initial recognition model, and testing the model classification accuracy by using the testing set data to obtain a target recognition model.
In the identification stage, the model is updated in the process of identifying the data to be identified, the information of the surface electromyographic signals to be identified is fused into the model parameters, and the classifier is continuously updated, and the process can be divided into two types: one is a supervised update and the other is an unsupervised update. The supervised updating is applicable to the case of a part of labeled data samples available for calibration, and updates model parameters by a supervised learning method by using the part of labeled calibration data; the non-supervision updating process uses a part of non-label data samples as calibration data, and updates model parameters by using an non-supervision learning mode. In the individual difference problem, the model can be more quickly adapted to the tested individual by using the supervised update, and the recognition rate of the gesture action is not reduced when the individual uses the model for a long time based on the unsupervised update.
It should be noted that, for simplicity of description, the foregoing method embodiments are all expressed as a series of action combinations, but it should be understood by those skilled in the art that the present application is not limited by the order of actions described, as some steps may be performed in other order or simultaneously in accordance with the present application. Further, those skilled in the art will also appreciate that the embodiments described in the specification are all preferred embodiments, and that the acts and modules referred to are not necessarily required in the present application.
From the description of the above embodiments, it will be clear to a person skilled in the art that the method according to the above embodiments may be implemented by means of software plus the necessary general hardware platform, but of course also by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk), comprising several instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method described in the embodiments of the present application.
According to another aspect of the embodiments of the present application, there is also provided an apparatus for updating an action pattern recognition model for implementing the method for updating an action pattern recognition model described above. FIG. 5 is a schematic diagram of an alternative device for updating an action pattern recognition model according to an embodiment of the present application, as shown in FIG. 5, the device may include:
a first obtaining module 52, configured to obtain a first surface electromyographic signal, where the first surface electromyographic signal is a surface electromyographic signal to be identified collected from a first object;
the recognition module 54 is configured to recognize the first surface electromyographic signal by using a first motion pattern recognition model, and obtain a recognition result, where the first motion pattern recognition model is obtained by training an initial motion pattern recognition model by using a second surface electromyographic signal, and the second surface electromyographic signal is a surface electromyographic signal sample collected from the first object;
a second obtaining module 56, configured to obtain, from the first surface electromyographic signal, a target surface electromyographic signal whose identification result meets a target condition;
a first updating module 58, configured to update the first motion pattern recognition model with the target surface electromyographic signal to obtain a second motion pattern recognition model, where the second motion pattern recognition model is used to perform motion pattern recognition on the first object.
It should be noted that, the first obtaining module 52 in this embodiment may be used to perform step S202 in the embodiment of the present application, the identifying module 54 in this embodiment may be used to perform step S204 in the embodiment of the present application, the second obtaining module 56 in this embodiment may be used to perform step S206 in the embodiment of the present application, and the first updating module 58 in this embodiment may be used to perform step S208 in the embodiment of the present application.
It should be noted that the above modules are the same as examples and application scenarios implemented by the corresponding steps, but are not limited to what is disclosed in the above embodiments. It should be noted that the above modules may be implemented in software or hardware as a part of the apparatus in the hardware environment shown in fig. 1.
Through the module, in the process of identifying the first surface electromyographic signals to be identified of the first object by the first action pattern identification model obtained through training, the target surface electromyographic signals with the identification result meeting the target condition are collected to update the first action pattern identification model to obtain the second action pattern identification model, the purpose of updating the model in real time by adopting an unsupervised mode in the identification process is achieved, the adaptability of the model to the first object is improved, the technical effect of improving the adaptability of the action pattern identification model is achieved, and the technical problem that the adaptability of the action pattern identification model in the related technology is poor is solved.
As an alternative embodiment, the identification module includes:
the identification unit is used for identifying the first surface electromyographic signals by using the first action mode identification model to obtain prediction labels corresponding to the first surface electromyographic signals;
and the first determining unit is used for determining the confidence coefficient of the predictive label, wherein the identification result comprises the predictive label and the confidence coefficient.
As an alternative embodiment, the second obtaining module includes:
a first obtaining unit, configured to obtain a third surface electromyographic signal from the first surface electromyographic signal, where the predicted tag corresponding to the third surface electromyographic signal is a tag with an identification error;
and a second determining unit configured to determine the third surface electromyographic signal as the target surface electromyographic signal.
As an alternative embodiment, the second obtaining module further includes:
the second acquisition unit is used for acquiring a fourth surface electromyographic signal from the first surface electromyographic signal, wherein the predicted label corresponding to the fourth surface electromyographic signal is a label with correct identification;
a third obtaining unit, configured to obtain a fifth surface electromyographic signal, where the confidence coefficient corresponding to the fifth surface electromyographic signal is lower than a target confidence coefficient, from the fourth surface electromyographic signal;
and a third determining unit configured to determine the fifth surface electromyographic signal as the target surface electromyographic signal.
As an alternative embodiment, the apparatus further comprises:
the third acquisition module is used for acquiring a sixth surface electromyographic signal after updating the first action pattern recognition model by using the target surface electromyographic signal to obtain a second action pattern recognition model, wherein the sixth surface electromyographic signal is a surface electromyographic signal sample acquired from a second object;
and the second updating module is used for updating the second action pattern recognition model by using the sixth surface electromyographic signal to obtain a third action pattern recognition model, wherein the third action pattern recognition model is used for carrying out action pattern recognition on the second object.
It should be noted that the above modules are the same as examples and application scenarios implemented by the corresponding steps, but are not limited to what is disclosed in the above embodiments. It should be noted that the above modules may be implemented in software or in hardware as part of the apparatus shown in fig. 1, where the hardware environment includes a network environment.
According to another aspect of the embodiments of the present application, there is also provided a server or a terminal for implementing the method for updating an action pattern recognition model described above.
Fig. 6 is a block diagram of a terminal according to an embodiment of the present application, and as shown in fig. 6, the terminal may include: one or more (only one is shown in the figure) processors 601, memory 603, and transmission means 605, as shown in fig. 6, the terminal may further comprise an input output device 607.
The memory 603 may be configured to store software programs and modules, such as program instructions/modules corresponding to the method and apparatus for updating an action pattern recognition model in the embodiments of the present application, and the processor 601 executes the software programs and modules stored in the memory 603, thereby performing various functional applications and data processing, that is, implementing the method for updating an action pattern recognition model described above. Memory 603 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid state memory. In some examples, the memory 603 may further include memory remotely located with respect to the processor 601, which may be connected to the terminal through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The transmission device 605 is used to receive or transmit data via a network, and may also be used for data transmission between the processor and the memory. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 605 includes a network adapter (Network Interface Controller, NIC) that may be connected to other network devices and routers via a network cable to communicate with the internet or a local area network. In one example, the transmission device 605 is a Radio Frequency (RF) module that is configured to communicate wirelessly with the internet.
In particular, the memory 603 is used to store applications.
The processor 601 may call an application program stored in the memory 603 through the transmission means 605 to perform the steps of:
s1, acquiring a first surface electromyographic signal, wherein the first surface electromyographic signal is a surface electromyographic signal to be identified, which is acquired from a first object;
s2, recognizing the first surface electromyographic signals by using a first action pattern recognition model to obtain a recognition result, wherein the first action pattern recognition model is obtained by training an initial action pattern recognition model by using a second surface electromyographic signal, and the second surface electromyographic signal is a surface electromyographic signal sample acquired from the first object;
s3, acquiring a target surface electromyographic signal, of which the identification result meets a target condition, from the first surface electromyographic signal;
and S4, updating the first action mode identification model by using the target surface electromyographic signals to obtain a second action mode identification model, wherein the second action mode identification model is used for carrying out action mode identification on the first object.
By adopting the embodiment of the application, an updating scheme of the action pattern recognition model is provided. In the process of identifying the first surface electromyographic signals to be identified of the first object through the first action pattern identification model obtained through training, the target surface electromyographic signals with the identification result meeting the target condition are collected to update the first action pattern identification model to obtain the second action pattern identification model, the purpose of updating the model in real time by adopting an unsupervised mode in the identification process is achieved, the adaptability of the model to the first object is improved, the technical effect of improving the adaptability of the action pattern identification model is achieved, and the technical problem that the adaptability of the action pattern identification model in the related technology is poor is solved.
Alternatively, specific examples in this embodiment may refer to examples described in the foregoing embodiments, and this embodiment is not described herein.
It will be appreciated by those skilled in the art that the structure shown in fig. 6 is only illustrative, and the terminal may be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a palmtop computer, a mobile internet device (Mobile Internet Devices, MID), a PAD, etc. Fig. 6 is not limited to the structure of the electronic device. For example, the terminal may also include more or fewer components (e.g., network interfaces, display devices, etc.) than shown in fig. 6, or have a different configuration than shown in fig. 6.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be implemented by a program for instructing a terminal device to execute in association with hardware, the program may be stored in a computer readable storage medium, and the storage medium may include: flash disk, read-Only Memory (ROM), random-access Memory (Random Access Memory, RAM), magnetic or optical disk, and the like.
Embodiments of the present application also provide a storage medium. Alternatively, in the present embodiment, the above-described storage medium may be used for program code for executing the update method of the action pattern recognition model.
Alternatively, in this embodiment, the storage medium may be located on at least one network device of the plurality of network devices in the network shown in the above embodiment.
Alternatively, in the present embodiment, the storage medium is configured to store program code for performing the steps of:
s1, acquiring a first surface electromyographic signal, wherein the first surface electromyographic signal is a surface electromyographic signal to be identified, which is acquired from a first object;
s2, recognizing the first surface electromyographic signals by using a first action pattern recognition model to obtain a recognition result, wherein the first action pattern recognition model is obtained by training an initial action pattern recognition model by using a second surface electromyographic signal, and the second surface electromyographic signal is a surface electromyographic signal sample acquired from the first object;
s3, acquiring a target surface electromyographic signal, of which the identification result meets a target condition, from the first surface electromyographic signal;
and S4, updating the first action mode identification model by using the target surface electromyographic signals to obtain a second action mode identification model, wherein the second action mode identification model is used for carrying out action mode identification on the first object.
Alternatively, specific examples in this embodiment may refer to examples described in the foregoing embodiments, and this embodiment is not described herein.
Alternatively, in the present embodiment, the storage medium may include, but is not limited to: a U-disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The foregoing embodiment numbers of the present application are merely for describing, and do not represent advantages or disadvantages of the embodiments.
The integrated units in the above embodiments may be stored in the above-described computer-readable storage medium if implemented in the form of software functional units and sold or used as separate products. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions to cause one or more computer devices (which may be personal computers, servers or network devices, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application.
In the foregoing embodiments of the present application, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
In several embodiments provided in the present application, it should be understood that the disclosed client may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, such as the division of the units, is merely a logical function division, and may be implemented in another manner, for example, multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The foregoing is merely a preferred embodiment of the present application and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present application and are intended to be comprehended within the scope of the present application.