Accelerated reconstruction of a compressively sampled data stream

Pantelis Sopasakis, Nikolaos Freris, Panagiotis Patrinos

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

Abstract

The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed: the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements. In this paper, we develop a novel Newton-type forwardbackward proximal method to recursively solve the regularized Least-Squares problem (LASSO) online. We establish global convergence of our method as well as a local quadratic convergence rate. Our simulations show a substantial speed-up over the state of the art which may render the proposed method suitable for applications with stringent real-time constraints.

Original languageEnglish (US)
Title of host publication2016 24th European Signal Processing Conference, EUSIPCO 2016
PublisherEuropean Signal Processing Conference, EUSIPCO
Pages1078-1082
Number of pages5
Volume2016-November
ISBN (Electronic)9780992862657
DOIs
StatePublished - Nov 28 2016
Event24th European Signal Processing Conference, EUSIPCO 2016 - Budapest, Hungary
Duration: Aug 28 2016Sep 2 2016

Other

Other24th European Signal Processing Conference, EUSIPCO 2016
CountryHungary
CityBudapest
Period8/28/169/2/16

Fingerprint

Compressed sensing
Sampling

Keywords

  • Compressed sensing
  • Forward backward splitting
  • LASSO
  • Operator splitting methods
  • Recursive algorithms

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Sopasakis, P., Freris, N., & Patrinos, P. (2016). Accelerated reconstruction of a compressively sampled data stream. In 2016 24th European Signal Processing Conference, EUSIPCO 2016 (Vol. 2016-November, pp. 1078-1082). [7760414] European Signal Processing Conference, EUSIPCO. https://doi.org/10.1109/EUSIPCO.2016.7760414

Accelerated reconstruction of a compressively sampled data stream. / Sopasakis, Pantelis; Freris, Nikolaos; Patrinos, Panagiotis.

2016 24th European Signal Processing Conference, EUSIPCO 2016. Vol. 2016-November European Signal Processing Conference, EUSIPCO, 2016. p. 1078-1082 7760414.

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

Sopasakis, P, Freris, N & Patrinos, P 2016, Accelerated reconstruction of a compressively sampled data stream. in 2016 24th European Signal Processing Conference, EUSIPCO 2016. vol. 2016-November, 7760414, European Signal Processing Conference, EUSIPCO, pp. 1078-1082, 24th European Signal Processing Conference, EUSIPCO 2016, Budapest, Hungary, 8/28/16. https://doi.org/10.1109/EUSIPCO.2016.7760414
Sopasakis P, Freris N, Patrinos P. Accelerated reconstruction of a compressively sampled data stream. In 2016 24th European Signal Processing Conference, EUSIPCO 2016. Vol. 2016-November. European Signal Processing Conference, EUSIPCO. 2016. p. 1078-1082. 7760414 https://doi.org/10.1109/EUSIPCO.2016.7760414
Sopasakis, Pantelis ; Freris, Nikolaos ; Patrinos, Panagiotis. / Accelerated reconstruction of a compressively sampled data stream. 2016 24th European Signal Processing Conference, EUSIPCO 2016. Vol. 2016-November European Signal Processing Conference, EUSIPCO, 2016. pp. 1078-1082
@inproceedings{258bba02329b4f04a72995fe03fd91a9,
title = "Accelerated reconstruction of a compressively sampled data stream",
abstract = "The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed: the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements. In this paper, we develop a novel Newton-type forwardbackward proximal method to recursively solve the regularized Least-Squares problem (LASSO) online. We establish global convergence of our method as well as a local quadratic convergence rate. Our simulations show a substantial speed-up over the state of the art which may render the proposed method suitable for applications with stringent real-time constraints.",
keywords = "Compressed sensing, Forward backward splitting, LASSO, Operator splitting methods, Recursive algorithms",
author = "Pantelis Sopasakis and Nikolaos Freris and Panagiotis Patrinos",
year = "2016",
month = "11",
day = "28",
doi = "10.1109/EUSIPCO.2016.7760414",
language = "English (US)",
volume = "2016-November",
pages = "1078--1082",
booktitle = "2016 24th European Signal Processing Conference, EUSIPCO 2016",
publisher = "European Signal Processing Conference, EUSIPCO",

}

TY - GEN

T1 - Accelerated reconstruction of a compressively sampled data stream

AU - Sopasakis, Pantelis

AU - Freris, Nikolaos

AU - Patrinos, Panagiotis

PY - 2016/11/28

Y1 - 2016/11/28

N2 - The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed: the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements. In this paper, we develop a novel Newton-type forwardbackward proximal method to recursively solve the regularized Least-Squares problem (LASSO) online. We establish global convergence of our method as well as a local quadratic convergence rate. Our simulations show a substantial speed-up over the state of the art which may render the proposed method suitable for applications with stringent real-time constraints.

AB - The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed: the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements. In this paper, we develop a novel Newton-type forwardbackward proximal method to recursively solve the regularized Least-Squares problem (LASSO) online. We establish global convergence of our method as well as a local quadratic convergence rate. Our simulations show a substantial speed-up over the state of the art which may render the proposed method suitable for applications with stringent real-time constraints.

KW - Compressed sensing

KW - Forward backward splitting

KW - LASSO

KW - Operator splitting methods

KW - Recursive algorithms

UR - http://www.scopus.com/inward/record.url?scp=85006049678&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006049678&partnerID=8YFLogxK

U2 - 10.1109/EUSIPCO.2016.7760414

DO - 10.1109/EUSIPCO.2016.7760414

M3 - Conference contribution

AN - SCOPUS:85006049678

VL - 2016-November

SP - 1078

EP - 1082

BT - 2016 24th European Signal Processing Conference, EUSIPCO 2016

PB - European Signal Processing Conference, EUSIPCO

ER -