Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Motion intention recognition method of EEG signals based on multitask RNN model

A technology of EEG signals and recognition methods, applied in character and pattern recognition, biological neural network models, graphic reading, etc., can solve problems such as signal noise and timing information that cannot be processed well

Active Publication Date: 2019-02-22
NORTHEAST NORMAL UNIVERSITY
View PDF3 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] In order to solve the defect that the prior art cannot process signal noise and timing information well, the present invention provides a method for recognizing action intentions of EEG signals based on a multi-task RNN ​​model

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
  • Motion intention recognition method of EEG signals based on multitask RNN model
  • Motion intention recognition method of EEG signals based on multitask RNN model
  • Motion intention recognition method of EEG signals based on multitask RNN model

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0037] Specific implementation mode 1. Combination Figure 1 to Figure 3 Describe this embodiment, the EEG signal action intention recognition method based on the multi-task RNN ​​model,

[0038]This embodiment proposes a novel framework in which a multi-task recurrent neural network learns to separate EEG signals. Since noise can only account for a large proportion at a specific frequency, we first decompose the EEG signal into different frequency channels to reduce the interference of noise from other frequency signals. In this way, more robust feature learning can be achieved compared to non-separated signal representation methods.

[0039] For each band channel, the signal with real information will be learned by the recurrent neural network to obtain a complex representation of this band. Specifically, a special type of recurrent neural network (RNN), long short-term memory network (LSTM), is used to forge a temporal feature from signals with different frequency bands. ...

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

Multi-task RNN model based EEG signal action intention recognitioin method is disclosed. The invention relates to a method for identifying the action intention of EEG signals by a multi-task recurrentneural network, and solves the defect that the prior art cannot well process the signal noise and the time sequence information. The invention provides a novel framework for learning the unique characteristics from the EEG signals by applying the multi-task recurrent neural network. The correlation between different EEG frequency bands and biological significance is improved by learning the separated signals used for human motion intention recognition. Temporal correlation between different channels is also exploited. In this way, the identification of binary and multivariate intentions is improved. The invention performs extensive experiments on publicly available EEG signal reference datasets and compares our method with many of the most advanced algorithms. The experimental results show that the method proposed by the invention surpasses all comparison methods and achieves an accuracy rate of 97.8%.

Description

technical field [0001] The invention relates to a method for recognizing action intentions of electroencephalogram signals by a multi-task recurrent neural network. Background technique [0002] The human brain is clearly a complex system, and brain-computer interfaces (BCIs) can convert neural activity into signals, allowing studies to discover correlations between brain activity and behavior. Electroencephalogram (EEG) signal analysis is a non-invasive technique to obtain brain activity through BCI, which can reflect the brain activity of subjects when performing specific tasks. [0003] Most of the intention recognition work based on EEG signals only represents the characteristics of EEG signals with a frequency. Such processing is unfavorable to the effect of the subsequent learning model. The information in an EEG signal segment can be further divided into different frequency ranges, where each frequency range has a different correlation level for a specific brain act...

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
IPC IPC(8): G06F3/01G06K9/00G06N3/04
CPCG06F3/015G06N3/048G06F2218/12
Inventor 岳琳陈炜通殷明浩赵晓威赵浩男
Owner NORTHEAST NORMAL UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com