BCI remote control

Abstract

Since the introduction of television remote control in the 1950s, the technologies brought great convenience to the users of various electronic products. However, the user's experience lingered around this button-based methodology ever since. The recent advancement in Brain-Computer Interface technology has shed light on the way human interacts with electronic devices. In particular, the Electroencephalography (EEG) technique enables the “reading” of human mind in a safe, convenient and affordable fashion. The aim of this project is to design an EEG based remote controller to replace the traditional button-based methods. To achieve this, various signal processing and machine learning techniques were applied to analyze and classify the raw data obtained from EEG sensors. Specifically, the EEG data is first transformed into frequency domain using Fast Fourier Transform (FFT), and then projected onto lower dimensions using Principal Component Analysis (PCA). Finally, the EEG data were classified with Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) classifiers. In this report, the traditional human-computer interface designs were compared with the EEG brain-computer interfacing method. Several signal processing and machine learning algorithms for EEG applications were briefly introduced. After that, the details of the system design were presented. Finally, the machine learning method's training and testing results were reported.