Brain 2 Communicate: EEG-based Affect Recognition to Augment Virtual Social Interactions

Abstract

The perception and expression of emotion is a fundamental

part of social interaction. This project aims to utilize neuronal

signals to augment avatar-mediated communications. We

recognize emotions with a brain-computer-interface (BCI)

and supervised machine learning. Using an avatar-based

communication interface that supports head tracking, gaze

tracking, and speech to animation, we leverage the BCI-based

affect detection to visualize emotional states.