Toward A Brain Interface For Tracking Attended Auditory Sources
Marzieh Haghighi, Northeastern University
Mohammad Moghadamfalahi, Northeastern University
Hooman Nezamfar, Northeastern University
Murat Akcakaya, Electrical, Computer Engineering Department University of Pittsburgh
Deniz Erdogmus, Northeastern University

Auditory-evoked noninvasive electroencephalography (EEG) based brain-computer interfaces (BCIs) could be useful for improved hearing aids in the future. This manuscript investigates the role of frequency and spatial features of audio signal in EEG activities in an auditory BCI system with the purpose of detecting the attended auditory source in a cocktail party setting. A cross correlation based feature between EEG and speech envelope is shown to be useful to discriminate attention in the case of two different speakers. Results indicate that, on average, for speaker and direction (of arrival) of audio signals classification, the presented approach yields 91% and 86% accuracy, respectively.