Analog neuromorphic computation: an application to compression

Fernando J Pineda, Andreas G. Andreou

Research output: Contribution to journalArticle

Abstract

Nature has evolved computing engines whose intelligence and natural abilities are unrivaled by modern computers. To match Mother Nature's abilities, we must overcome the same difficulties faced by natural systems, and we must learn to perform reliable computing with unreliable components. Steps in this direction are being taken by several groups at the Applied Physics Laboratory and at The Johns Hopkins University Department of Electrical and Computer Engineering. The purpose of the work is twofold: (1) to explore algorithms based on physical and neural models of computation and (2) to develop useful applications. We describe the basic approach and an experimental electronic neural network for the decompression of one-dimensional signals.

Original languageEnglish (US)
Pages (from-to)82-85
Number of pages4
JournalJohns Hopkins APL Technical Digest (Applied Physics Laboratory)
Volume15
Issue number1
StatePublished - Jan 1994
Externally publishedYes

Fingerprint

analogs
electrical engineering
intelligence
pressure reduction
engines
Physics
engineering
Engines
Neural networks
physics
electronics

ASJC Scopus subject areas

  • General
  • Physics and Astronomy (miscellaneous)

Cite this

Analog neuromorphic computation : an application to compression. / Pineda, Fernando J; Andreou, Andreas G.

In: Johns Hopkins APL Technical Digest (Applied Physics Laboratory), Vol. 15, No. 1, 01.1994, p. 82-85.

Research output: Contribution to journalArticle

@article{29e9284599dd41da89fb70587bf7efc5,
title = "Analog neuromorphic computation: an application to compression",
abstract = "Nature has evolved computing engines whose intelligence and natural abilities are unrivaled by modern computers. To match Mother Nature's abilities, we must overcome the same difficulties faced by natural systems, and we must learn to perform reliable computing with unreliable components. Steps in this direction are being taken by several groups at the Applied Physics Laboratory and at The Johns Hopkins University Department of Electrical and Computer Engineering. The purpose of the work is twofold: (1) to explore algorithms based on physical and neural models of computation and (2) to develop useful applications. We describe the basic approach and an experimental electronic neural network for the decompression of one-dimensional signals.",
author = "Pineda, {Fernando J} and Andreou, {Andreas G.}",
year = "1994",
month = "1",
language = "English (US)",
volume = "15",
pages = "82--85",
journal = "Johns Hopkins APL Technical Digest (Applied Physics Laboratory)",
issn = "0270-5214",
publisher = "John Hopkins University Press",
number = "1",

}

TY - JOUR

T1 - Analog neuromorphic computation

T2 - an application to compression

AU - Pineda, Fernando J

AU - Andreou, Andreas G.

PY - 1994/1

Y1 - 1994/1

N2 - Nature has evolved computing engines whose intelligence and natural abilities are unrivaled by modern computers. To match Mother Nature's abilities, we must overcome the same difficulties faced by natural systems, and we must learn to perform reliable computing with unreliable components. Steps in this direction are being taken by several groups at the Applied Physics Laboratory and at The Johns Hopkins University Department of Electrical and Computer Engineering. The purpose of the work is twofold: (1) to explore algorithms based on physical and neural models of computation and (2) to develop useful applications. We describe the basic approach and an experimental electronic neural network for the decompression of one-dimensional signals.

AB - Nature has evolved computing engines whose intelligence and natural abilities are unrivaled by modern computers. To match Mother Nature's abilities, we must overcome the same difficulties faced by natural systems, and we must learn to perform reliable computing with unreliable components. Steps in this direction are being taken by several groups at the Applied Physics Laboratory and at The Johns Hopkins University Department of Electrical and Computer Engineering. The purpose of the work is twofold: (1) to explore algorithms based on physical and neural models of computation and (2) to develop useful applications. We describe the basic approach and an experimental electronic neural network for the decompression of one-dimensional signals.

UR - http://www.scopus.com/inward/record.url?scp=0028207782&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028207782&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0028207782

VL - 15

SP - 82

EP - 85

JO - Johns Hopkins APL Technical Digest (Applied Physics Laboratory)

JF - Johns Hopkins APL Technical Digest (Applied Physics Laboratory)

SN - 0270-5214

IS - 1

ER -