Overall Objectives
Scientific Foundations
Application Domains
New Results
Contracts and Grants with Industry
Other Grants and Activities

Section: New Results

Keywords : source separation, sparse representation, adaptive basis, source localization, probabilistic source model.

Source separation

Source separation via sparse and adaptive representations

Participants : Emmanuel Vincent, Remi Gribonval.

Main collaboration: Andrew Nesbit (Queen Mary, University of London), Matthieu Puigt (Laboratoire d'Astrophysique de Toulouse-Tarbes)

Source separation is the task of retrieving the source signals underlying a multichannel mixture signal, where each channel is the sum of filtered versions of the sources. The state-of-the-art approach consists of representing the signals in a given time-frequency basis and estimating the source coefficients by sparse decomposition in that basis, under an exact mixture reconstruction constraint relying on a frequency-wise approximation of the mixing process. This approach often provides limited performance due to poor approximation of the mixing process in reverberant environments and to the use of a time-frequency basis where the sources overlap. Our previous work on adaptive stereo bases [18] showed promising results but suggested that the modeling of the mixing process and the choice of an adapted basis should be separately addressed so as to avoid over-fitting issues. We investigated the replacement of the mixture reconstruction constraint by a quadratic penalty term computed from the true mixing process, resulting in improved separation performance in reverberant environments with large microphone spacing [26] . We also studied a range of adaptive lapped orthogonal time-frequency bases originally designed for audio coding and explained how to estimate the best basis in a source separation context [35] , [50] , [49] . Finally, we provided an experimental validation of the implicit source independence assumption underlying the above approaches [51] .

A new probabilistic framework for source separation

Participants : Simon Arberet, Remi Gribonval, Emmanuel Vincent, Frédéric Bimbot.

Main collaboration: Alexey Ozerov (Telecom ParisTech)

In parallel with our work on sparse representations, we proposed a new framework for audio source separation where each source is modeled as a zero-mean Gaussian variable in the neighborhood of each time-frequency bin. This framework was first applied to the problem of source counting and localization and resulted in increased robustness by selection of the time-frequency bins with a single active source [46] . We subsequently investigated its use for the problem of source separation by defining two distinct models for the source variances: either a mild sparsity prior in each time-frequency bin [54] or a GMM prior introducing some dependencies between the variances in different frequency bins [48] . Both approaches were tested over instantaneous mixtures and provided respectively a significant improvement of the separation performance over all mixtures and an even larger improvement over music mixtures.


Logo Inria