Subject: IEEE Special Issue Call for Papers From: Paris Smaragdis <paris(at)MEDIA.MIT.EDU> Date: Wed, 8 Dec 2004 14:01:50 -0500Dear friends, this is an announcement of special issue from the IEEE transactions on speech and signal processing which we think you will find interesting. You can find more details on the issue's web page at: http://www.sapa2004.org/IEEESAP/ Submission deadline is January 31 2005. Paris IEEE Transactions on Speech and Audio Processing Special Issue on Statistical and Perceptual Audio Processing http://www.sapa2004.org/IEEESAP/ Systems for automatic audio analysis fall into two broad camps: those based on statistical methods and models (e.g. current speech recognizers, independent component analysis), and those mimicking perceptual systems (exemplified by computational auditory scene analysis, as well as early approaches to speech recognition such as blackboards). Both these approaches have their merits, and their strengths appear complementary. The goal of this special issue is to collect work that integrates these two threads - using rigorous, statistical tools to exploit cues and processing analogous to those used by listeners. In this special issue we invite researchers to submit papers that combine perceptual and statistical principles, as applied to speech, music and audio analysis. Papers describing original and previously unpublished relevant research and new concepts are solicited on, but not limited to, the following topics: * General audio analysis * Speech analysis * Music analysis * Audio classification * Speech recognition * Signal separation * Multi-channel analysis * Computational Auditory Scene Analysis (CASA) * Perceptual aspects of statistical algorithms such as independent component analysis and non-negative matrix factorization. * Hybrid methods that use CASA-like cues in a statistical framework, e.g. for separation or recognition. * Theoretical and empirical results on the unification of statistical and perceptually based approaches.