Description of the project
FASST is a Flexible Audio Source Separation Toolbox in Matlab, designed to speed up the conception and automate the implementation of new model-based audio source separation algorithms. Based on a local-Gaussian modelling of sound sources and exploiting spatial and spectral cues on the sources to separate, it allows the flexible specifications of constraints or degrees of freedom of different nature in a unified framework.
FASST is currently available in Matlab. Porting to C++ has been started in January 2013 and should be achieved by the end of 2014.