Joint sparse recovery based on variances

Ben Adcock, Anne Gelb, Guohui Song, Yi Sui

Research output: Contribution to journalArticle

Abstract

Much research has recently been devoted to sparse signal recovery and image reconstruction from multiple measurement vectors. The assumption that the underlying signals or images have some common features with sparse representation suggests that using a joint sparsity approach to recover each individual signal or image can be more effective than recovering each signal or image separately using standard sparse recovery techniques. Joint sparsity reconstruction is typically performed using \ell 2,1-minimization, and although the process yields better results than separate recoveries, the inherent coupling makes the algorithm computationally intensive, since it is difficult to parallelize. In this investigation, we first observe that the elementwise variance of the signals convey information about their shared support. This observation motivates us to introduce a weighted \ell 1-joint sparsity algorithm, where the weights depend on the calculated variance. Specifically, the \ell 1-minimization should be more heavily penalized in regions where the corresponding variance is small, since it is likely there is no signal there. We demonstrate that our new method, which we refer to as variance-based joint sparse recovery, is more accurate and cost efficient. Applications in sparse signal recovery, parallel magnetic resonance imaging, and edge detection are considered.

Original languageEnglish (US)
Pages (from-to)A246-A268
JournalSIAM Journal on Scientific Computing
Volume41
Issue number1
DOIs
StatePublished - Jan 1 2019
Externally publishedYes

Fingerprint

Recovery
Sparsity
Edge detection
Magnetic resonance
Image reconstruction
Sparse Representation
Magnetic Resonance Imaging
Edge Detection
Image Reconstruction
Imaging techniques
Likely
Costs
Demonstrate

Keywords

  • Joint sparsity
  • Multiple measurement vectors
  • Sparse signal recovery
  • Variance

ASJC Scopus subject areas

  • Computational Mathematics
  • Applied Mathematics

Cite this

Joint sparse recovery based on variances. / Adcock, Ben; Gelb, Anne; Song, Guohui; Sui, Yi.

In: SIAM Journal on Scientific Computing, Vol. 41, No. 1, 01.01.2019, p. A246-A268.

Research output: Contribution to journalArticle

Adcock, Ben ; Gelb, Anne ; Song, Guohui ; Sui, Yi. / Joint sparse recovery based on variances. In: SIAM Journal on Scientific Computing. 2019 ; Vol. 41, No. 1. pp. A246-A268.
@article{453895d9312e471da5aac44217ccff80,
title = "Joint sparse recovery based on variances",
abstract = "Much research has recently been devoted to sparse signal recovery and image reconstruction from multiple measurement vectors. The assumption that the underlying signals or images have some common features with sparse representation suggests that using a joint sparsity approach to recover each individual signal or image can be more effective than recovering each signal or image separately using standard sparse recovery techniques. Joint sparsity reconstruction is typically performed using \ell 2,1-minimization, and although the process yields better results than separate recoveries, the inherent coupling makes the algorithm computationally intensive, since it is difficult to parallelize. In this investigation, we first observe that the elementwise variance of the signals convey information about their shared support. This observation motivates us to introduce a weighted \ell 1-joint sparsity algorithm, where the weights depend on the calculated variance. Specifically, the \ell 1-minimization should be more heavily penalized in regions where the corresponding variance is small, since it is likely there is no signal there. We demonstrate that our new method, which we refer to as variance-based joint sparse recovery, is more accurate and cost efficient. Applications in sparse signal recovery, parallel magnetic resonance imaging, and edge detection are considered.",
keywords = "Joint sparsity, Multiple measurement vectors, Sparse signal recovery, Variance",
author = "Ben Adcock and Anne Gelb and Guohui Song and Yi Sui",
year = "2019",
month = "1",
day = "1",
doi = "10.1137/17M1155983",
language = "English (US)",
volume = "41",
pages = "A246--A268",
journal = "SIAM Journal of Scientific Computing",
issn = "1064-8275",
publisher = "Society for Industrial and Applied Mathematics Publications",
number = "1",

}

TY - JOUR

T1 - Joint sparse recovery based on variances

AU - Adcock, Ben

AU - Gelb, Anne

AU - Song, Guohui

AU - Sui, Yi

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Much research has recently been devoted to sparse signal recovery and image reconstruction from multiple measurement vectors. The assumption that the underlying signals or images have some common features with sparse representation suggests that using a joint sparsity approach to recover each individual signal or image can be more effective than recovering each signal or image separately using standard sparse recovery techniques. Joint sparsity reconstruction is typically performed using \ell 2,1-minimization, and although the process yields better results than separate recoveries, the inherent coupling makes the algorithm computationally intensive, since it is difficult to parallelize. In this investigation, we first observe that the elementwise variance of the signals convey information about their shared support. This observation motivates us to introduce a weighted \ell 1-joint sparsity algorithm, where the weights depend on the calculated variance. Specifically, the \ell 1-minimization should be more heavily penalized in regions where the corresponding variance is small, since it is likely there is no signal there. We demonstrate that our new method, which we refer to as variance-based joint sparse recovery, is more accurate and cost efficient. Applications in sparse signal recovery, parallel magnetic resonance imaging, and edge detection are considered.

AB - Much research has recently been devoted to sparse signal recovery and image reconstruction from multiple measurement vectors. The assumption that the underlying signals or images have some common features with sparse representation suggests that using a joint sparsity approach to recover each individual signal or image can be more effective than recovering each signal or image separately using standard sparse recovery techniques. Joint sparsity reconstruction is typically performed using \ell 2,1-minimization, and although the process yields better results than separate recoveries, the inherent coupling makes the algorithm computationally intensive, since it is difficult to parallelize. In this investigation, we first observe that the elementwise variance of the signals convey information about their shared support. This observation motivates us to introduce a weighted \ell 1-joint sparsity algorithm, where the weights depend on the calculated variance. Specifically, the \ell 1-minimization should be more heavily penalized in regions where the corresponding variance is small, since it is likely there is no signal there. We demonstrate that our new method, which we refer to as variance-based joint sparse recovery, is more accurate and cost efficient. Applications in sparse signal recovery, parallel magnetic resonance imaging, and edge detection are considered.

KW - Joint sparsity

KW - Multiple measurement vectors

KW - Sparse signal recovery

KW - Variance

UR - http://www.scopus.com/inward/record.url?scp=85063040505&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063040505&partnerID=8YFLogxK

U2 - 10.1137/17M1155983

DO - 10.1137/17M1155983

M3 - Article

VL - 41

SP - A246-A268

JO - SIAM Journal of Scientific Computing

JF - SIAM Journal of Scientific Computing

SN - 1064-8275

IS - 1

ER -