AES New York 2013
Workshop W11
Friday, October 18, 10:30 am — 12:30 pm (Room 1E08)
W11 - Audio Source Separation
Chair:Gautham Mysore, Adobe Research - San Francisco, CA, USA
Panelists:
Nicholas Bryan, Stanford University - Stanford, CA, USA
Derry FitzGerald, Cork Institute of Technology - Cork, Ireland
Elias Kokkinis, accusonus - Patras, Greece
Pierre Leveau, Audionamix - Paris, France
Abstract:
Audio source separation algorithms aim to take a recording of a mixture of sound sources as an input and provide the separated sources as outputs. Algorithmically, this is a very challenging problem. However, some recent technological advances have made this possible for multiple real world scenarios such as denoising in the presence of complex noises, pitch correcting certain notes while preserving others, processing only the vocals of a song while preserving the background music, extracting dialogue from old films to provide a higher quality soundtrack, removing microphone leakage from multichannel drum recordings, upmixing mono to stereo with panning of sound sources, and more generally, music remixing. Some of these technologies are available in products (Adobe Audition CC, Melodyne, ISSE, ADX Trax, Drumatom). Others are used by specialized sound engineers and are offered as a service (Audionamix, Derry Fitzgerald). This panel is comprised of some of the inventors of these technologies, who will discuss the ideas and their practical use.
 | This session is presented in association with the AES Technical Committee on Semantic Audio Analysis |
