Subject: International Conference on Pattern Recognition From: Stephen Smoliar <smoliar(at)ISS.NUS.SG> Date: Sat, 11 Sep 1993 02:27:18 -0400Al, I hope you are enjoying Palo Alto. I was there last month wishing I could stay longer! I have been working on material for a paper for the International Conference on Pattern Recognition on and off for most of this calendar year. I even brought a bunch of data on the PowerBook which accompanied me on my trip, and I am now banging out a draft. My working title is "In Search of Musical Objects." I am interested in the extent to which one can identify points in the temporal stream which may be interpreted as boundaries between "objects" when one is listening to music. I chose, as my point of departure, some use Peter Desain has made of autocorrelation in analyzing the structure of expressive performance of rhythms. I have been applying his technique to both score and performance data in a variety of ways and have come up with some results which seem to jive rather nicely with a Lerdahl-Jackendoff-like metrical structure analysis. Is this likely to be of interest? Steve