Graduate Research Talk: Using fNIRS for Implicit Interfaces

Abstract

A Brain-Computer Interface (BCI) siphons ‘free' information from the user; it describes the user’s cognitive/emotional state or intentions without demanding that he or she press a button or move a mouse. This talk explores BCIs that use functional near-infrared spectroscopy (fNIRS) as implicit input. I describe how effective research work requires dedicated work in cognitive / neuroscience, data mining, and user interface design. As cognitive scientists, we form plausible estimates about what fNIRS-detectable brain regions underlie what mental functions. As software engineers, we write programs that translate measurements of oxygenation changes into predictions of the user’s state. As designers, we explore what state’s could be relevant to a user interface, and how to handle error prone input.