Learning Sparse Codes for Image Reconstruction

Kai Labusch, Thomas Martinetz


We propose a new algorithm for the design of overcomplete dictionaries for sparse coding that generalizes the Sparse Coding Neural Gas (SCNG) algorithm such that it is not bound to a particular approximation method for the coecients of the dictionary elements. In an application to image reconstruction, a dictionary that has been learned using this algorithm outperforms a dictionary that has been obtained from the widely-used K-SVD algorithm, an overcomplete Haar-wavelet dictionary and an overcomplete discrete cosine transformation (DCT).
Original languageEnglish
Title of host publicationProceedings of the 18th European Symposium on Artificial Neural Networks
EditorsMichel Verleysen
Number of pages6
Publication date04.2010
ISBN (Print)2-930307-10-2
Publication statusPublished - 04.2010
Event18th European Symposium on Artificial Neural Networks - Bruges, Belgium
Duration: 28.04.201030.04.2010


Dive into the research topics of 'Learning Sparse Codes for Image Reconstruction'. Together they form a unique fingerprint.

Cite this