In recent years many source separation algorithms which are based on different types of second order statistics have been proposed. Among these we find the SOBI, the phase-SCORE and many others. We describe a unifying frame under which all these seemingly different algorithms can be combined into a single coherent algebraic source separation framework. We also derive a new whitening scheme which greatly enhances each of the algorithms in the family, and a new generalization of the phase-SCORE algorithm.
|Title of host publication||Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics, SPW-HOS 1999|
|Publisher||Institute of Electrical and Electronics Engineers Inc.|
|Number of pages||5|
|ISBN (Electronic)||0769501400, 9780769501406|
|State||Published - 1999|
|Event||1999 IEEE Signal Processing Workshop on Higher-Order Statistics, SPW-HOS 1999 - Caesarea, Israel|
Duration: 14 Jun 1999 → 16 Jun 1999
|Name||Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics, SPW-HOS 1999|
|Conference||1999 IEEE Signal Processing Workshop on Higher-Order Statistics, SPW-HOS 1999|
|Period||14/06/99 → 16/06/99|
Bibliographical noteFunding Information:
Amir Leshem was supported by the NOEMI project of the STW under contract no. DEL77-4476. Email: firstname.lastname@example.org.
© 1999 IEEE.
- Source separation
- bilinear forms
- generalized second order statistics