000087115 001__ 87115
000087115 005__ 20190316233747.0
000087115 02470 $$2ISI$$a000224752800124
000087115 037__ $$aCONF
000087115 245__ $$aSparse decomposition over multi-component redundant dictionaries
000087115 269__ $$a2004
000087115 260__ $$bIEEE$$c2004
000087115 336__ $$aConference Papers
000087115 490__ $$aDatabases and Information Systems
000087115 520__ $$aIn many applications - such as compression, de-noising and source separation - a good and efficient signal representation is characterized by sparsity. This means that many coefficients are close to zero, while only few ones have a non-negligible amplitude. On the other hand, real-world signals - such as audio or natural images - clearly present peculiar structures. In this paper we introduce a global optimization framework that aims at respecting the sparsity criterion while decomposing a signal over an overcomplete, multi-component dictionary. We adopt a probabilistic analysis which can lead to consider the signal internal structure. As an example that fits this framework, we propose the Weighted Basis Pursuit algorithm, based on the solution of a convex, non-quadratic problem. Results show that this method can provide sparse signal representations and sparse m-terms approximations. Moreover, Weighted Basis Pursuit provides a faster convergence compared to Basis Pursuit.
000087115 6531_ $$aLTS2
000087115 700__ $$0241529$$g141038$$aGranai, L.
000087115 700__ $$aVandergheynst, P.$$g120906$$0240428
000087115 773__ $$tMultimedia Signal Processing (MMSP04), Workshop on$$q494-497
000087115 8564_ $$uhttps://infoscience.epfl.ch/record/87115/files/Granai2004_864.pdf$$zn/a$$s109671
000087115 909C0 $$xU10380$$0252392$$pLTS2
000087115 909CO $$qGLOBAL_SET$$pconf$$ooai:infoscience.tind.io:87115$$pSTI
000087115 937__ $$aEPFL-CONF-87115
000087115 970__ $$aGranai2004_864/LTS
000087115 973__ $$sPUBLISHED$$aEPFL
000087115 980__ $$aCONF