CCABSS - Blind Source Separation by Canonical Correlation Analysis

Y = CCABSS(X) is the BSS of X=A*S where S is a set of unknown source signals

and A is an unknown mixing matrix. The components in S are supposed to

be independent. Y is an estimate of S appart from permutation and scaling.

For mixed 1-D signals, X is 2-D. The first index refer to the different

components and the second index refers to the signal parameter (e.g. time)

For mixed images, X is 3-D where the first index refers to the different

mixed images and the second and third indeces are the spatial coordinates.



[Y W] = CCABSS(X) also gives the de-mixing matrix W, such that Y = W *X.-CCABSS -　Blind Source Separation by Canonical Correlation Analysis



　Y = CCABSS(X) is the BSS of X=A*S where S is a set of unknown source signals

　and A is an unknown mixing matrix. The components in S are supposed to

　be independent. Y is an estimate of S appart from permutation and scaling.

　For mixed 1-D signals, X is 2-D. The first index refer to the different

　components and the second index refers to the signal parameter (e.g. time)

　For mixed images, X is 3-D where the first index refers to the different

　mixed images and the second and third indeces are the spatial coordinates.



　[Y W] = CCABSS(X) also gives the　de-mixing　matrix W, such that Y = W *X.