SOURCE SEPARATION IN THE PRESENCE OF SIDE INFORMATION: NECESSARY AND SUFFICIENT CONDITIONS FOR RELIABLE DE-MIXING

This paper puts forth new recovery guarantees for the source separation problem in the presence of side information, where one observes the linear superposition of two source signals plus two additional signals that are correlated with the mixed ones. By positing that the individual components of the mixed signals as well as the corresponding side information signals follow a joint Gaussian mixture model, we characterise necessary and sufficient conditions for reliable separation in the asymptotic regime of low-noise as a function of the geometry of the underlying signals and their interaction. In particular, we show that if the subspaces spanned by the innovation components of the source signals with respect to the side information signals have zero intersection, provided that we observe a certain number of measurements from the mixture, then we can reliably separate the sources, otherwise we cannot. We also provide a number of numerical results on synthetic data that validate our theoretical findings.

[1]  Mohamed-Jalal Fadili,et al.  Sparsity and Morphological Diversity in Blind Source Separation , 2007, IEEE Transactions on Image Processing.

[2]  Barak A. Pearlmutter,et al.  Blind Source Separation by Sparse Decomposition in a Signal Dictionary , 2001, Neural Computation.

[3]  Miguel R. D. Rodrigues,et al.  Compressed Sensing With Prior Information: Strategies, Geometry, and Bounds , 2017, IEEE Transactions on Information Theory.

[4]  Maïté Brandt-Pearce.  Signal separation using fractional sampling in multiuser communications , 2000, IEEE Trans. Commun..

[5]  C. Yau,et al.  Bayesian non‐parametric hidden Markov models with applications in genomics , 2011 .

[6]  Stanislaw Gorlow,et al.  Informed separation of spatial images of stereo music recordings using second-order statistics , 2013, 2013 IEEE International Workshop on Machine Learning for Signal Processing (MLSP).

[7]  Laurent Girin,et al.  A Watermarking-Based Method for Informed Source Separation of Audio Signals With a Single Sensor , 2010, IEEE Transactions on Audio, Speech, and Language Processing.

[8]  Stéphane Mallat,et al.  Solving Inverse Problems With Piecewise Linear Estimators: From Gaussian Mixture Models to Structured Sparsity , 2010, IEEE Transactions on Image Processing.

[9]  A. Robert Calderbank,et al.  Classification and Reconstruction of High-Dimensional Signals From Low-Dimensional Features in the Presence of Side Information , 2014, IEEE Transactions on Information Theory.

[10]  Karim Abed-Meraim,et al.  Underdetermined Blind Audio Source Separation Using Modal Decomposition , 2007, EURASIP J. Audio Speech Music. Process..

[11]  Erkki Oja,et al.  Independent component analysis: algorithms and applications , 2000, Neural Networks.

[12]  Miguel R. D. Rodrigues,et al.  Multi-Modal Dictionary Learning for Image Separation With Application in Art Investigation , 2016, IEEE Transactions on Image Processing.

[13]  Guillermo Sapiro,et al.  Statistical Compressed Sensing of Gaussian Mixture Models , 2011, IEEE Transactions on Signal Processing.

[14]  Guillermo Sapiro,et al.  Video Compressive Sensing Using Gaussian Mixture Models , 2014, IEEE Transactions on Image Processing.

[15]  A. Robert Calderbank,et al.  Reconstruction of Signals Drawn From a Gaussian Mixture Via Noisy Compressive Measurements , 2013, IEEE Transactions on Signal Processing.

[16]  Antoine Liutkus,et al.  Informed source separation through spectrogram coding and data embedding , 2012, Signal Process..

[17]  David B. Dunson,et al.  Compressive Sensing on Manifolds Using a Nonparametric Mixture of Factor Analyzers: Algorithm and Performance Bounds , 2010, IEEE Transactions on Signal Processing.

[18]  A. Robert Calderbank,et al.  Communications Inspired Linear Discriminant Analysis , 2012, ICML.