Lefschetz Center for Dynamical Systems Seminar
Brown Applied Mathematics Pattern Theory and Vision Seminar
Abstract:
I discuss my efforts to build an intelligent musical accompanist. I
divide the problem into three basic pieces. The "listening" piece
tries to track the score position of the soloist by using a hidden
Markov model. The "think" piece tries to combine the information
gained from "listening" with data gained from past performances. I do
this by estimating parameters for a Kalman filter model and using the
"on line" data to predict the soloist's actions. A final piece
mediates between these predictions and internal consistency
constraints the accompaniment must satisfy. A brief demonstration
will follow.
**** PLEASE NOTE NEW DAY AND TIME FROM LAST TERM ****
Scientific Computing Seminar
<--- 1997 Index