Efficient image processing via compressive sensing of integrate-and-fire neuronal network dynamics

Victor J. Barranca, Gregor Kovačič, Doug Zhou, David Cai

Research output: Contribution to journalArticle

Abstract

Integrate-and-fire (I&F) neuronal networks are ubiquitous in diverse image processing applications, including image segmentation and visual perception. While conventional I&F network image processing requires the number of nodes composing the network to be equal to the number of image pixels driving the network, we determine whether I&F dynamics can accurately transmit image information when there are significantly fewer nodes than network input-signal components. Although compressive sensing (CS) theory facilitates the recovery of images using very few samples through linear signal processing, it does not address whether similar signal recovery techniques facilitate reconstructions through measurement of the nonlinear dynamics of an I&F network. In this paper, we present a new framework for recovering sparse inputs of nonlinear neuronal networks via compressive sensing. By recovering both one-dimensional inputs and two-dimensional images, resembling natural stimuli, we demonstrate that input information can be well-preserved through nonlinear I&F network dynamics even when the number of network-output measurements is significantly smaller than the number of input-signal components. This work suggests an important extension of CS theory potentially useful in improving the processing of medical or natural images through I&F network dynamics and understanding the transmission of stimulus information across the visual system.

Original languageEnglish (US)
Pages (from-to)1313-1322
Number of pages10
JournalNeurocomputing
Volume171
DOIs
StatePublished - Jan 1 2016

Fingerprint

Visual Perception
Nonlinear Dynamics
Fires
Image processing
Recovery
Image segmentation
Signal processing
Pixels
Processing

Keywords

  • Compressive sensing
  • Neuronal networks
  • Nonlinear dynamics
  • Signal processing

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Cognitive Neuroscience

Cite this

Efficient image processing via compressive sensing of integrate-and-fire neuronal network dynamics. / Barranca, Victor J.; Kovačič, Gregor; Zhou, Doug; Cai, David.

In: Neurocomputing, Vol. 171, 01.01.2016, p. 1313-1322.

Research output: Contribution to journalArticle

Barranca, Victor J. ; Kovačič, Gregor ; Zhou, Doug ; Cai, David. / Efficient image processing via compressive sensing of integrate-and-fire neuronal network dynamics. In: Neurocomputing. 2016 ; Vol. 171. pp. 1313-1322.
@article{112a8a15ad5a47ec9e7fa8369737e0ff,
title = "Efficient image processing via compressive sensing of integrate-and-fire neuronal network dynamics",
abstract = "Integrate-and-fire (I&F) neuronal networks are ubiquitous in diverse image processing applications, including image segmentation and visual perception. While conventional I&F network image processing requires the number of nodes composing the network to be equal to the number of image pixels driving the network, we determine whether I&F dynamics can accurately transmit image information when there are significantly fewer nodes than network input-signal components. Although compressive sensing (CS) theory facilitates the recovery of images using very few samples through linear signal processing, it does not address whether similar signal recovery techniques facilitate reconstructions through measurement of the nonlinear dynamics of an I&F network. In this paper, we present a new framework for recovering sparse inputs of nonlinear neuronal networks via compressive sensing. By recovering both one-dimensional inputs and two-dimensional images, resembling natural stimuli, we demonstrate that input information can be well-preserved through nonlinear I&F network dynamics even when the number of network-output measurements is significantly smaller than the number of input-signal components. This work suggests an important extension of CS theory potentially useful in improving the processing of medical or natural images through I&F network dynamics and understanding the transmission of stimulus information across the visual system.",
keywords = "Compressive sensing, Neuronal networks, Nonlinear dynamics, Signal processing",
author = "Barranca, {Victor J.} and Gregor Kovačič and Doug Zhou and David Cai",
year = "2016",
month = "1",
day = "1",
doi = "10.1016/j.neucom.2015.07.067",
language = "English (US)",
volume = "171",
pages = "1313--1322",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",

}

TY - JOUR

T1 - Efficient image processing via compressive sensing of integrate-and-fire neuronal network dynamics

AU - Barranca, Victor J.

AU - Kovačič, Gregor

AU - Zhou, Doug

AU - Cai, David

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Integrate-and-fire (I&F) neuronal networks are ubiquitous in diverse image processing applications, including image segmentation and visual perception. While conventional I&F network image processing requires the number of nodes composing the network to be equal to the number of image pixels driving the network, we determine whether I&F dynamics can accurately transmit image information when there are significantly fewer nodes than network input-signal components. Although compressive sensing (CS) theory facilitates the recovery of images using very few samples through linear signal processing, it does not address whether similar signal recovery techniques facilitate reconstructions through measurement of the nonlinear dynamics of an I&F network. In this paper, we present a new framework for recovering sparse inputs of nonlinear neuronal networks via compressive sensing. By recovering both one-dimensional inputs and two-dimensional images, resembling natural stimuli, we demonstrate that input information can be well-preserved through nonlinear I&F network dynamics even when the number of network-output measurements is significantly smaller than the number of input-signal components. This work suggests an important extension of CS theory potentially useful in improving the processing of medical or natural images through I&F network dynamics and understanding the transmission of stimulus information across the visual system.

AB - Integrate-and-fire (I&F) neuronal networks are ubiquitous in diverse image processing applications, including image segmentation and visual perception. While conventional I&F network image processing requires the number of nodes composing the network to be equal to the number of image pixels driving the network, we determine whether I&F dynamics can accurately transmit image information when there are significantly fewer nodes than network input-signal components. Although compressive sensing (CS) theory facilitates the recovery of images using very few samples through linear signal processing, it does not address whether similar signal recovery techniques facilitate reconstructions through measurement of the nonlinear dynamics of an I&F network. In this paper, we present a new framework for recovering sparse inputs of nonlinear neuronal networks via compressive sensing. By recovering both one-dimensional inputs and two-dimensional images, resembling natural stimuli, we demonstrate that input information can be well-preserved through nonlinear I&F network dynamics even when the number of network-output measurements is significantly smaller than the number of input-signal components. This work suggests an important extension of CS theory potentially useful in improving the processing of medical or natural images through I&F network dynamics and understanding the transmission of stimulus information across the visual system.

KW - Compressive sensing

KW - Neuronal networks

KW - Nonlinear dynamics

KW - Signal processing

UR - http://www.scopus.com/inward/record.url?scp=84944516150&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84944516150&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2015.07.067

DO - 10.1016/j.neucom.2015.07.067

M3 - Article

AN - SCOPUS:84944516150

VL - 171

SP - 1313

EP - 1322

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

ER -