Schnass, KarinVandergheynst, Pierre2009-02-172009-02-172009-02-172011https://infoscience.epfl.ch/handle/20.500.14299/35331In this article we present a signal model for classification based on a collection of low dimensional subspaces embedded into the high dimensional signal space. We develop an alternate projection algorithm to find such a collection and finally test the classification performance of our scheme in comparison to Fisher's LDA and a recent approach based on sparse approximation.classificationGrassmannian manifoldssubspace learningdictionary learningalternate projectionsLTS2lts2text::journal::journal article::research article