The Flexible Audio Source Separation Toolbox Version 2.0

Abstract : The Flexible Audio Source Separation Toolbox (FASST) is a toolbox for audio source separation that relies on a general modeling and estimation framework that is applicable to a wide range of scenarios. We introduce the new version of the toolbox written in C++, which provides a number of advantages compared to the first Matlab version: portability, faster computation, simplified user interface, more scripting languages. In addition, we provide a state-of-the-art example of use for the separation of speech and domestic noise. The demonstration will give attendees the opportunity to explore the settings and to experience their effect on the separation performance.
Complete list of metadatas

Cited literature [9 references]  Display  Hide  Download

https://hal.inria.fr/hal-00957412
Contributor : Yann Salaün <>
Submitted on : Monday, March 10, 2014 - 4:13:06 PM
Last modification on : Saturday, March 30, 2019 - 1:26:31 AM
Long-term archiving on : Tuesday, June 10, 2014 - 12:01:34 PM

File

show-and-tell-ICASSP.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00957412, version 1

Citation

Yann Salaün, Emmanuel Vincent, Nancy Bertin, Nathan Souviraà-Labastie, Xabier Jaureguiberry, et al.. The Flexible Audio Source Separation Toolbox Version 2.0. ICASSP, May 2014, Florence, Italy. 2014. ⟨hal-00957412⟩

Share

Metrics

Record views

4841

Files downloads

2970