Towards sign language recognition using EEG-based motor imagery brain computer interface

Abstract

While BCIs have a wide range of applications, the majority of research in the field is concentrated on addressing the issues of controlling and communicating for paralysed patients. This research seeks to examine—through the completion of offline experimentation—a particular aspect; that is, the likelihood of linguistic communication with those paralysed patients, merely by means of neural activity in the brain. Electroencephalogram (EEG) brain activities obtained whilst imagining execution of six one-handed signs from American Sign Language (ASL) were investigated. Upon reviewing the findings, it is demonstrated that EEG signal analysis can be used efficiently to identify hand movement of sign language from the brain. SVM and LDA both showed the highest accuracy, achieving around 75% correct when the Entropy feature type was examined.