Chasing butterflies: In search of efficient dictionaries - Archive ouverte HAL Access content directly
Conference Papers Year : 2015

Chasing butterflies: In search of efficient dictionaries

Abstract

Dictionary learning aims at finding a frame (called dictionary) in which training data admits a sparse representation. Traditional dictionary learning is limited to relatively small-scale problems, because dense matrices associated to high-dimensional dictionaries can be costly to manipulate, both at the learning stage and when using this dictionary for tasks such as sparse coding. In this paper, inspired by usual fast transforms, we consider a dictionary structure allowing cheaper manipulation, and we propose a learning algorithm imposing this structure. The approach is demonstrated experimentally with the factorization of the Hadamard matrix and on image denoising.
Fichier principal
Vignette du fichier
ICASSP_current.pdf (970.65 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01104696 , version 1 (19-01-2015)
hal-01104696 , version 2 (23-01-2015)

Identifiers

Cite

Luc Le Magoarou, Rémi Gribonval. Chasing butterflies: In search of efficient dictionaries. International Conference on Acoustics, Speech and Signal Processing (ICASSP), Apr 2015, Brisbane, Australia. ⟨10.1109/ICASSP.2015.7178579⟩. ⟨hal-01104696v2⟩
1007 View
973 Download

Altmetric

Share

Gmail Facebook X LinkedIn More