Sohrob Kazerounian joined the Department of Cognitive and Neural Systems at Boston University in 2005 after finding out about the research goals of CELEST. As an undergraduate at the University of Connecticut, he earned dual degrees in Computer Science and Engineering and in Cognitive Science, with a minor in Mathematics. This combination of degrees allowed him to pursue his interests in the interdisciplinary fields of philosophy of mind, cognitive linguistics, and theory of computation.
Since joining CELEST, he has been working with Stephen Grossberg on developing a computational model of human speech perception and language acquisition. By simulating the underlying neural networks involved in speech perception, the model shows how bottom-up and top-down interactions give rise to the ability to perceive speech when portions of it are occluded by noise, and sometimes even absent from the acoustic signal altogether. The model explains this by showing how low-level auditory features are processed, grouped, stored in working memory, and learned as sequence chunks which can be committed to long-term memory.
After his involvement with the inter-Science of Learning Centers (iSLC) conferences in 2008 and 2009, which brought together students and postdocs from the six NSF Science of Learning Centers, Sohrob became a co-organizer of the 2010 iSLC conference. He is a past co-President of the CELEST student/postdoc organization, and was also one of the co-founders of CNSO, the Cognitive and Neural Systems Department student organization.
Last updated on May 18, 2011