Unlock instant, AI-driven research and patent intelligence for your innovation.

Surface electromyogram signal classification method based on space attention pruning capsule network

A technology of EMG signal and classification method, applied in the field of surface EMG signal classification based on spatial attention pruning capsule network, can solve the problem of increasing network attention weight, reducing computing cost, and not being able to obtain feature space correlation well and other issues to achieve good sparsity and reduce memory requirements

Pending Publication Date: 2022-07-01
ZHEJIANG UNIV CITY COLLEGE
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the shortcomings of the existing technical solutions that cannot well obtain the spatial correlation between features, the present invention proposes a surface electromyography signal classification method based on the spatial attention pruning capsule network, and the discrete features are calculated by Cartesian product operation. After two-dimensionalization, it is sent to the capsule network based on spatial attention for training. At the same time, by increasing the regularization loss to constrain the spatial saliency attention weight to increase the attention weight of the network to the central area, and through the dynamic pruning operation Remove low-level capsules that do not contribute much to high-level capsules to simplify the model and reduce computational costs

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Surface electromyogram signal classification method based on space attention pruning capsule network
  • Surface electromyogram signal classification method based on space attention pruning capsule network
  • Surface electromyogram signal classification method based on space attention pruning capsule network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036] like figure 1 As shown, the present invention provides a surface EMG signal classification method based on spatial attention pruning capsule network. The spatial attention information is obtained by convolution calculation, and then sent to the capsule network for training, and the regularization and pruning mechanism are added to increase its attention to the central feature, improve the classification accuracy and reduce the training and prediction time.

[0037] Further, the classification method specifically includes the following steps:

[0038] Step 1: For the surface EMG signals measured in all channels, use the window analysis method to process the surface EMG signals obtained by the recording electrodes; where w represents the window length, t represents the incremental interval, and τ represents the processing of feature extraction and classification operations Delay; after the interval of each time t, sequentially extract the features of the signal with time l...

example

[0069] Example: The surface EMG signals used in the study were collected by the ELONXI EMG acquisition instrument developed by a team at the University of Portsmouth, UK. The device supports up to 16 bipolar channels with a sampling resolution of 24 bits and a sampling frequency between 1000Hz and 2000Hz. The experimental dataset contains a total of 8 subjects' surface EMG signals measured in 6 different time periods. In each time period, each subject demonstrated 5 gestures, and the surface EMG signals of each gesture were detected and recorded by 16 bipolar electrodes (channels). To exclude the transition state between the two gestures, the middle 10 s of the surface EMG signal of each gesture action was marked as the steady-state signal. The sampling frequency of the surface EMG signal is set to 1kHz, then the original surface EMG data size of each gesture action is 10000×16. The original surface EMG signal size of the dataset is 8×6×5×10000×16, that is, 240×10000×16. Nu...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a surface electromyogram signal classification method based on a space attention pruning capsule network, and the method comprises the steps: collecting surface electromyogram signals, carrying out the preprocessing of the surface electromyogram signals, and obtaining target electromyogram signals; performing two-dimensional processing on the discrete features of the target electromyographic signals based on Cartesian product operation to obtain a two-dimensional feature relation graph; obtaining space attention information based on the two-dimensional characteristic relation graph; and constructing a spatial attention pruning capsule network model, and inputting the two-dimensional features into the spatial attention pruning capsule network model for feature classification to obtain a classification result. According to the method, the recognition precision of the model is improved, the training time is shortened, the short delay requirement of surface electromyogram signal gesture action recognition in the actual situation is met, and the method can adapt to sample recognition with a large data scale.

Description

technical field [0001] The invention belongs to the field of signal classification, in particular to a surface electromyographic signal classification method based on spatial attention pruning capsule network. Background technique [0002] When human beings entered the Internet era from the PC era, the mobile Internet has brought us massive explosions of information, medical data and other multi-attribute text data have grown rapidly, and the contradiction between information overload and lack of knowledge has become increasingly prominent. Therefore, how to efficiently deal with massive multi-attribute information has become the focus of current research. However, in the face of exponentially growing discrete feature attribute data, pure manual labeling and classification is unrealistic, so it is imminent to realize machine intelligent classification. [0003] Large-scale multi-classification datasets generally consist of multiple discrete feature attributes. These discre...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06N3/04G06N3/08A61B5/00A61B5/397
CPCG06N3/082A61B5/397A61B5/7264G06N3/045G06F2218/08G06F2218/12
Inventor 王铮赵燕伟陈国棋
Owner ZHEJIANG UNIV CITY COLLEGE