Neural Nets

Structure, Musical [Str]  >  Formalism  ]

Overcoming inter-processor communication bottlenecks for parallel computing systems has demanded a fundamental reappraisal of the ways in which such engines are put together. One solution which has stimulated particular interest in the field of computer music since the late 1980s has been the construction of massively parallel systems from hundreds or even thousands of very simple microprocessors which are sometimes little more than basic logic gates, with the object of developing models which have the capacity to emulate the neural networks that operate within the human brain. Such systems facilitate the development of systems which explore aspects of artificial intelligence, of particular value in the development of more intuitive interfaces for composer-machine communications as well as powerful and versatile computing engines for the synthesis and processing of compositional material. A notable example of such a massively parallel resource which has been used for music applications is the Connection Machine, built at MIT in 1987-8. (Source - Peter Manning (1993). Electronic and Computer Music, Second Edition. Oxford University Press.)

 

See also:

AI, Connectionism

 

Bibliography:

English - Español - Français - Deutch - Italiano

Alphabetical order - Chronological order

Morrison, Kenneth (1999). Toward the Creation of an {Intelligent} Machine Improvisor
Myatt, Tony (2002). Strategies for Interaction in {Construction 3}
Purwinds, Hendrik, Blankertz, Benjamin, Obermayer, Klaus (2000). Computing Auditory Perception
Rodet, Xavier (1991). What Would We Like to See Our Music Machines Capable of Doing?

Top