Online Control of a Virtual Object with Collaborative SSVEP

Abstract

In this paper, we propose a brain-computer interface (BCI) based on collaborative steady-state visually evoked potential (SSVEP). A technique for estimating the common direction of the gaze of multiple subjects is studied with a view to controlling a virtual object in a virtual environment. The electro-encephalograms (EEG) of eight volunteers are simultaneously recorded with two virtual cubes as visual stimuli. These two virtual cubes flicker at different rates, 6 Hz and 8 Hz, and the corresponding SSVEP is observed around the occipital area. The amplitude spectra of the EEG activity of individual subjects are analyzed, averaged, and synthesized to obtain the collaborative SSVEP. Machine learning is applied to estimate the common gaze direction of the eight subjects with the supervised data from fewer than eight subjects. The estimation accuracy is perfect only in the case of the collaborative SSVEP. One-dimensional control of a virtual ball is performed by controlling the common eye gaze direction, which induces the collaborative SSVEP.