Blind source separation of more sources than mixtures using overcomplete representations

By Te-Won Lee, Michael S. Lewicki, Mark A. Girolami, Terrence J. Sejnowski. 1999

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Brain Computer Interfaces, a Review (2012, background): “The number of sources is usually assumed to be less than or equal to the number of channels although more generalized ICA methods are possible [159].”
Learning Overcomplete Representations (2000, background): “Nonetheless, is it possible to separate three speakers on two channels with good ﬁdelity (Lee, Lewicki, Girolami, & Sejnowski, 1999).”
Accelerated Canonical Polyadic Decomposition Using Mode Reduction (2012, unknown): “The CP-SMBSS is very useful if some a priori knowledge about component diversities, such as independence [12], [13], nonnegativity [14]–[16], sparsity [17], etc, in at least one mode is available, which allows us to recover them from their linear mixtures via standard BSS algorithms.”
On the approximate W-disjoint orthogonality of speech (2002, background): “Other work which leverages properties of sources which are similar in spirit to approximate W-disjoint orthogonality include [3, 4], which considered time domain disjoint sources, and [5], which exploited the sparsity of the short-time Fourier transform when applied to speech and music signals.”
Underdetermined blind source separation based on sparse representation (2006, methodology): “In several references [9], [10], the mixing matrix and sources were estimated using the maximum posterior approach or the maximum-likelihood approach.”
Electrocortical activity is coupled to gait cycle phase during treadmill walking (2011, unknown): “…mixture independent component analysis (ICA) algorithm [AMICA] (Palmer et al., 2006, 2008) that generalizes infomax (Bell and Sejnowski, 1995; Lee et al., 1999a) and multiple mixtures (Lee et al., 1999b; Lewicki and Sejnowski, 2000) ICA approaches, to parse EEG signals into spatially static,…”
Underdetermined blind source separation using sparse representations (2001, background): “A similar two-step approach can be found in [8,9], where a learning rule for A is derived by #tting a multivariate Gaussian around the current estimate of the source components.”
Level of participation in robotic-assisted treadmill walking modulates midline sensorimotor EEG rhythms in able-bodied subjects (2012, methodology): “AMICA is a generalization of the Infomax algorithm (Bell and Sejnowski, 1995; Makeig et al., 1996) and multiple mixture (Lee et al., 1999; Lewicki and Sejnowski, 2000) ICA approaches.”
Blind separation of speech mixtures via time-frequency masking (2004, methodology): “In [2], [8], [11], and [12], the demixing is done by making additional assumptions on the statistical properties of the sources and using a maximum a posteriori (MAP) estimator.”
Analysis of Sparse Representation and Blind Source Separation (2004, methodology): “In several references, the mixing matrix and sources were estimated using the maximum posterior approach and the maximum likelihood approach (Zibulevsky & Pearlmutter, 2001; Zibulevsky et al., 2000; Lee, Lewicki, Girolami, & Sejnowski, 1999 ).”
Handbook of Blind Source Separation: Independent Component Analysis and Applications (2010, background): “Some ten years elapsed before these techniques began to be fully exploited for blind source separation [72, 106, 66, 70, 116, 61, 15, 111].”
Sparse component analysis and blind source separation of underdetermined mixtures (2005, background): “There is a large amount of papers devoted to ICA problems [2], [5] but mostly for thecase . We refer to [1], [6], [ 7 ], and [9]–[11] for some recent papers on SCA and underdetermined ICA .”