Intention estimation via support vector machines (SVM) and convolutional neuron networks (CNN) of brain activity signals

Abstract

Electroencephalography (EEG) and functional near infra-red spectroscopy (fNIRS) are popular noninvasive methods for brain-computer interfaces (BCI), due to the simplicity and portability of the systems. Although the systems have gone through great progress in development, a satisfiable intention estimation has not yet been developed, and is being tackled by many neural activity researchers. This paper will utilize EEG and FNIRS to estimate the intentions of its users, whether the user is resting, or moving their right or left hand via CNN and SVM.