Florian Blaettler
Email:
|
|
Work phone:
|
044/635'30'65
|
Home page:
|
|
Underdetermined Blind Source Separation:
The separation of different signal source which where mixed up
in a single (or in few) channel is still a widely unsolved
problem with potential applications such as speech recognition,
telecomunication or medical signal processing. The human brain
performs this kind of task with high reliability for the
auditory system (cocktail party effect).
The main objective of the project is the development and
implementation of a practical BSS algorithm. It has to cope
problems with and without prior knowledge of the original
signal. Currently non-negative independent component analysis is
considered as core of such an algorithm. Finally we want to run
it on real-life environment and compare the output to
measurements of brain function.