Australian Journal of Intelligent Information Processing Systems, Vol 11, No 3 (2010): Computational Neuroscience and Brain Computer Interface

Font Size:  Small  Medium  Large

Mind Flipper: An EEG-Based Brain Computer Interface for Page-Turning During Presentation

Yunjun Nam, Hyohyeong Kang, Qibin Zhao, Andrzej Cichocki, Seungjin Choi

Abstract


In this paper we present a system called mind flipper which is an EEG-based brain computer interface (BCI), enabling a computer to automatically turn a slide forward during an oral presentation by analyzing brain waves involving imagery or real movements of a subject’s right hand. False positive errors are critical in this application. To reduce false positive errors, we develop a nonparametric classifier based on convex hulls. Compared to existing linear binary classifiers, we show that the convex hull-based classifier easily removes ambiguous regions on a feature space determined by common spatial patterns, which result in false positive and negative errors. Experiments confirmed the useful behavior of our convex hull-based classifier, compared to linear or RBF-kernel based SVM.

Full Text: PDF