000117531 001__ 117531
000117531 005__ 20190316234134.0
000117531 037__ $$aCONF
000117531 245__ $$aDictionary Identifiability from Few Training Samples
000117531 260__ $$c2008
000117531 269__ $$a2008
000117531 336__ $$aConference Papers
000117531 520__ $$aThis article treats the problem of learning a dictionary providing sparse representations for a given signal class, via $ell^1$ minimisation, or more precisely the problem of identifying a dictionary $dico$ from a set of training samples $Y$ knowing that $Y = dico X$ for some coefficient matrix $X$. It provides a characterisation of coefficient matrices $X$ that allow to recover any orthonormal basis (ONB) as a local minimum of an $ell^1$ minimisation problem. Based on this characterisation it is shown that certain types of sparse random coefficient matrices will ensure local identifiability of the ONB with high probability.
000117531 6531_ $$alts2
000117531 6531_ $$asparse representation
000117531 6531_ $$adictionary learning
000117531 6531_ $$abasis learning
000117531 6531_ $$arecovery condition
000117531 6531_ $$a random coefficients
000117531 700__ $$aGribonval, Remi
000117531 700__ $$0240457$$g168927$$aSchnass, Karin
000117531 7112_ $$dAugust 2008$$cLausanne$$aEUSIPCO
000117531 773__ $$tProc. EUSIPCO'08
000117531 8564_ $$zURL
000117531 8564_ $$uhttps://infoscience.epfl.ch/record/117531/files/eusipco08.pdf$$zn/a$$s140321
000117531 909C0 $$xU10380$$0252392$$pLTS2
000117531 909CO $$ooai:infoscience.tind.io:117531$$qGLOBAL_SET$$pconf$$pSTI
000117531 937__ $$aEPFL-CONF-117531
000117531 973__ $$rREVIEWED$$sPUBLISHED$$aEPFL
000117531 980__ $$aCONF