Beyond the Narrowband Approximation: Wideband Convex Methods for Under-Determined Reverberant Audio Source Separation
Authors:
- 10 May 2010
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech, and Language Processing
- Vol. 18 (7) , 1818-1829
Abstract
We consider the problem of extracting the source signals from an under-determined convolutive mixture assuming known mixing filters. State-of-the-art methods operate in the time-frequency domain and rely on narrowband approximation of the convolutive mixing process by complex-valued multiplication in each frequency bin. The source signals are then estimated by minimizing either a mixture fitting cost or a ℓ1 source sparsity cost, under possible constraints on the number of active sources. In this paper, we define a wideband ℓ2 mixture fitting cost circumventing the above approximation and investigate the use of a ℓ1,2 mixed-norm cost promoting disjointness of the source time-frequency representations. We design a family of convex functionals combining these costs and derive suitable optimization algorithms. Experiments indicate that the proposed wideband methods result in a signal-to-distortion ratio improvement of 2 to 5 dB compared to the state-of-the-art on reverberant speech mixtures.Keywords
This publication has 23 references indexed in Scilit:
- Sparse regression using mixed normsApplied and Computational Harmonic Analysis, 2009
- On the performance of algorithms for the minimization of ℓ 1 -penalized functionalsInverse Problems, 2009
- Fixed-Point Continuation for $\ell_1$-Minimization: Methodology and ConvergenceSIAM Journal on Optimization, 2008
- Oracle estimators for the benchmarking of source separation algorithmsSignal Processing, 2007
- MAP-Based Underdetermined Blind Source Separation of Convolutive Mixtures by Hierarchical Clustering and -Norm MinimizationEURASIP Journal on Advances in Signal Processing, 2006
- Signal Recovery by Proximal Forward-Backward SplittingMultiscale Modeling & Simulation, 2005
- An iterative thresholding algorithm for linear inverse problems with a sparsity constraintCommunications on Pure and Applied Mathematics, 2004
- Blind Separation of Speech Mixtures via Time-Frequency MaskingIEEE Transactions on Signal Processing, 2004
- An affine scaling methodology for best basis selectionIEEE Transactions on Signal Processing, 1999
- Atomic Decomposition by Basis PursuitSIAM Journal on Scientific Computing, 1998