Learning multilevel dictionaries for compressed sensing using discriminative clustering

Jayaraman J. Thiagarajan, Karthikeyan Natesan Ramamurthy, Andreas Spanias, Panos Nasiopoulos

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The performance of sparse recovery using compressed measurements improves when dictionaries learned from training data are used in place of predefined dictionaries. In this paper, we propose to learn incoherent multilevel dictionaries using discriminative clustering in each level. To this end, we present the discriminative K-lines clustering that iterates between identifying the cluster centers and computing the discriminant directions. A scheme for computing representations using the proposed dictionary is also developed. Simulation results for compressed sensing using standard images demonstrate that incorporating incoherence in the dictionary results in improved recovery performance. Furthermore, we implement the proposed algorithms as part of a sparse representations toolbox for the J-DSP software package.

Original languageEnglish (US)
Title of host publicationProceedings of the 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012
PublisherIEEE Computer Society
Pages494-497
Number of pages4
ISBN (Print)9780769547121
DOIs
StatePublished - Jan 1 2012
Event2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012 - Piraeus-Athens, Greece
Duration: Jul 18 2012Jul 20 2012

Publication series

NameProceedings of the 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012

Other

Other2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012
CountryGreece
CityPiraeus-Athens
Period7/18/127/20/12

Keywords

  • Discriminative clustering
  • compressed sensing
  • incoherent dictionaries
  • sparse representations

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition
  • Signal Processing

Fingerprint Dive into the research topics of 'Learning multilevel dictionaries for compressed sensing using discriminative clustering'. Together they form a unique fingerprint.

  • Cite this

    Thiagarajan, J. J., Ramamurthy, K. N., Spanias, A., & Nasiopoulos, P. (2012). Learning multilevel dictionaries for compressed sensing using discriminative clustering. In Proceedings of the 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012 (pp. 494-497). [6274289] (Proceedings of the 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012). IEEE Computer Society. https://doi.org/10.1109/IIH-MSP.2012.125