Compressed sensing using a gaussian scale mixtures model in wavelet domain

Yookyung Kim, Mariappan S. Nadar, Ali Bilgin

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

33 Citations (Scopus)

Abstract

Compressed Sensing (CS) theory has gained attention recently as an alternative to the current paradigm of sampling followed by compression. Early CS recovery techniques operated under the implicit assumption that the transform coefficients in the sparsity domain are independently distributed. Recent works, however, demonstrated that exploiting the statistical dependencies between transform coefficients can further improve the recovery performance of CS. In this paper, we propose the use of a Gaussian Scale Mixtures (GSM) model in CS. This model can efficiently exploit the statistical dependencies between wavelet coefficients during CS recovery. The proposed model is incorporated into several recent CS techniques including Reweighted l1 minimization (RL1), Iteratively Reweighted Least Squares (IRLS), and Iterative Hard Thresholding (IHT). Experimental results show that the proposed method improves reconstruction quality for a given number of measurements or requires fewer measurements for a desired reconstruction quality.

Original languageEnglish (US)
Title of host publicationProceedings - International Conference on Image Processing, ICIP
Pages3365-3368
Number of pages4
DOIs
StatePublished - 2010
Event2010 17th IEEE International Conference on Image Processing, ICIP 2010 - Hong Kong, Hong Kong
Duration: Sep 26 2010Sep 29 2010

Other

Other2010 17th IEEE International Conference on Image Processing, ICIP 2010
CountryHong Kong
CityHong Kong
Period9/26/109/29/10

Fingerprint

Compressed sensing
Recovery
Sampling

Keywords

  • Compressed sensing
  • Gaussian scale mixtures
  • Wavelets

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition
  • Signal Processing

Cite this

Kim, Y., Nadar, M. S., & Bilgin, A. (2010). Compressed sensing using a gaussian scale mixtures model in wavelet domain. In Proceedings - International Conference on Image Processing, ICIP (pp. 3365-3368). [5652744] https://doi.org/10.1109/ICIP.2010.5652744

Compressed sensing using a gaussian scale mixtures model in wavelet domain. / Kim, Yookyung; Nadar, Mariappan S.; Bilgin, Ali.

Proceedings - International Conference on Image Processing, ICIP. 2010. p. 3365-3368 5652744.

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

Kim, Y, Nadar, MS & Bilgin, A 2010, Compressed sensing using a gaussian scale mixtures model in wavelet domain. in Proceedings - International Conference on Image Processing, ICIP., 5652744, pp. 3365-3368, 2010 17th IEEE International Conference on Image Processing, ICIP 2010, Hong Kong, Hong Kong, 9/26/10. https://doi.org/10.1109/ICIP.2010.5652744
Kim Y, Nadar MS, Bilgin A. Compressed sensing using a gaussian scale mixtures model in wavelet domain. In Proceedings - International Conference on Image Processing, ICIP. 2010. p. 3365-3368. 5652744 https://doi.org/10.1109/ICIP.2010.5652744
Kim, Yookyung ; Nadar, Mariappan S. ; Bilgin, Ali. / Compressed sensing using a gaussian scale mixtures model in wavelet domain. Proceedings - International Conference on Image Processing, ICIP. 2010. pp. 3365-3368
@inproceedings{d8bc57790c20423ca28c15d08c65992b,
title = "Compressed sensing using a gaussian scale mixtures model in wavelet domain",
abstract = "Compressed Sensing (CS) theory has gained attention recently as an alternative to the current paradigm of sampling followed by compression. Early CS recovery techniques operated under the implicit assumption that the transform coefficients in the sparsity domain are independently distributed. Recent works, however, demonstrated that exploiting the statistical dependencies between transform coefficients can further improve the recovery performance of CS. In this paper, we propose the use of a Gaussian Scale Mixtures (GSM) model in CS. This model can efficiently exploit the statistical dependencies between wavelet coefficients during CS recovery. The proposed model is incorporated into several recent CS techniques including Reweighted l1 minimization (RL1), Iteratively Reweighted Least Squares (IRLS), and Iterative Hard Thresholding (IHT). Experimental results show that the proposed method improves reconstruction quality for a given number of measurements or requires fewer measurements for a desired reconstruction quality.",
keywords = "Compressed sensing, Gaussian scale mixtures, Wavelets",
author = "Yookyung Kim and Nadar, {Mariappan S.} and Ali Bilgin",
year = "2010",
doi = "10.1109/ICIP.2010.5652744",
language = "English (US)",
isbn = "9781424479948",
pages = "3365--3368",
booktitle = "Proceedings - International Conference on Image Processing, ICIP",

}

TY - GEN

T1 - Compressed sensing using a gaussian scale mixtures model in wavelet domain

AU - Kim, Yookyung

AU - Nadar, Mariappan S.

AU - Bilgin, Ali

PY - 2010

Y1 - 2010

N2 - Compressed Sensing (CS) theory has gained attention recently as an alternative to the current paradigm of sampling followed by compression. Early CS recovery techniques operated under the implicit assumption that the transform coefficients in the sparsity domain are independently distributed. Recent works, however, demonstrated that exploiting the statistical dependencies between transform coefficients can further improve the recovery performance of CS. In this paper, we propose the use of a Gaussian Scale Mixtures (GSM) model in CS. This model can efficiently exploit the statistical dependencies between wavelet coefficients during CS recovery. The proposed model is incorporated into several recent CS techniques including Reweighted l1 minimization (RL1), Iteratively Reweighted Least Squares (IRLS), and Iterative Hard Thresholding (IHT). Experimental results show that the proposed method improves reconstruction quality for a given number of measurements or requires fewer measurements for a desired reconstruction quality.

AB - Compressed Sensing (CS) theory has gained attention recently as an alternative to the current paradigm of sampling followed by compression. Early CS recovery techniques operated under the implicit assumption that the transform coefficients in the sparsity domain are independently distributed. Recent works, however, demonstrated that exploiting the statistical dependencies between transform coefficients can further improve the recovery performance of CS. In this paper, we propose the use of a Gaussian Scale Mixtures (GSM) model in CS. This model can efficiently exploit the statistical dependencies between wavelet coefficients during CS recovery. The proposed model is incorporated into several recent CS techniques including Reweighted l1 minimization (RL1), Iteratively Reweighted Least Squares (IRLS), and Iterative Hard Thresholding (IHT). Experimental results show that the proposed method improves reconstruction quality for a given number of measurements or requires fewer measurements for a desired reconstruction quality.

KW - Compressed sensing

KW - Gaussian scale mixtures

KW - Wavelets

UR - http://www.scopus.com/inward/record.url?scp=78651087356&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78651087356&partnerID=8YFLogxK

U2 - 10.1109/ICIP.2010.5652744

DO - 10.1109/ICIP.2010.5652744

M3 - Conference contribution

AN - SCOPUS:78651087356

SN - 9781424479948

SP - 3365

EP - 3368

BT - Proceedings - International Conference on Image Processing, ICIP

ER -