Events
The Taub Faculty of Computer Science Events and Talks
Yotam Michael (EE, Technion)
Tuesday, 17.07.2012, 11:30
Blind Source Separation (BSS) is a very applicable and well-studied problem. Most studies of the BSS problem
assume the system to be time/position invariant, an assumption which assists the mathematical study,
but is not guaranteed for real world situations. We present a method applicable to the case of
underdetermined time/position varying mixing systems, where number of mixture observation is limited to
be less than the number of sources and the system is changing with time/position according to a parametric model. Our method is based on Staged Sparse Component Analysis (SSCA), which uses signal sparseness to estimate the varying system and then inverse it to allow source estimation. A variation called
Underdetermined SSCA (UDSSCA) is defined, which uses the signal sparseness for signal estimation
instead of mixing system inversion. The process of signal estimation is performed in a sparse domain
by the use of a winner-takes-all heuristic strategy inspired by minimization of l1 norm