Date: Wed, 20 Nov 1996 19:16:35 GMT Server: Apache/1.0.3 Content-type: text/html Content-length: 2585 Last-modified: Mon, 08 Jul 1996 15:14:23 GMT
Music and spoken language appear to be organized around hierarchically structured patterns of strong and weak accents - that is, they have rhythm. People are expert at recognizing the rhythmic structure in music and speech and guiding their own musical and linguistic behavior to match (that is, be entrained to) a perceived rhythm. Yet this capacity is not well understood, and no computer program even approaches it. Our project is concerned (1) with experimental investigation of the perception and production of rhythm in people and (2) with the development of a computational model of the learning and processing of simple rhythmic patterns. The evolving model is based on the idea that periodicities in the world must be captured by periodicities in the activity of the computational device which is responsible for dealing with it.In more general terms, this project is concerned with methods for establishment of a `temporal lock' between a computational device and the environment, that is, with entrainment between the realtime activity of a computer and the world. We suspect that such temporal locking is essential for human-like speech and music behavior -- and probably many other behaviors as well: catching a ball, dribbling a basketball, pushing a child on a swing, and so forth.
Associated Faculty: Robert Port, Michael Gasser
Associated Postdoc: Mauri Kaipainen (1996-7)
Associated Graduate Students: Doug Eck, Paul Kienzle, Fred Cummins (Linguistics)
Affiliated Projects: Accent Reduction Project: Diane Kewley-Port, Robert Port
Support: Office of Naval Research
For more information click here