SuperSID Final Report  Exploiting High-level Information for High-performance Speaker Recognition
				SuperSID Team WS02 21 August 2002