Research of Asynchronous Imitating-Reading BCI Based on Extreme Learning Machine
LI Yang
XIE Guodong
GUAN Jin'an
Abstract：The imitating-reading brain computer interface works on synchronous mode,while the asynchronous brain comput?er interface enables the user to control the system autonomously and has more practical value.In order to achieve asynchrony,work?ing state and the idle state need to be identified effectively.In this paper,using the extreme learning machine classification method based on the typical single-hidden layer feed-forward neural network to discuss the characteristics selection and mode classification of working state in imitating-reading mode. The target stimulus or non-target stimulus is used as working state feature,classified with three idle states,such as eyes closed,freedom thought,text reading and its hybrid form.The results show that the target stimu?lus time period as the working state feature can be distinguished effectively from the idle state,and the non-target stimulus time pe?riod as the working state feature can not be distinguished effectively.The hybrid form of three idle states has stable recognition accu?racy and good universality.
Keywords：imitating-readingasynchronous BCIidle stateextreme learning machine
Publication Date：2018-01-01
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：6( 479-484 )