Natural learning of neural networks by reconfiguration

L. Spaanenburg; R. Alberts; Cornelis H. Slump; B.J. van der Zwaag

doi:10.1117/12.499549

Natural learning of neural networks by reconfiguration

L. Spaanenburg, R. Alberts, Cornelis H. Slump, B.J. van der Zwaag

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

3 Citations (Scopus)

143 Downloads (Pure)

Abstract

The communicational and computational demands of neural networks are hard to satisfy in a digital technology. Temporal computing addresses this problem by iteration, but leaves a slow network. Spatial computing only became an option with the coming of modern FPGA devices. The paper provides two examples. First the balance between area and time is discussed on the realization of a modular feed-forward network. Second, the design of real-time image processing through a Cellular Neural Network is treated. In both examples, reconfiguration can be applied to provide for a natural and transparent support of learning.

Original language	Undefined
Title of host publication	Bioengineered and Bioinspired Systems
Editors	Angel Rodriguez-Vazquez, Derek Abbott, Ricardo Carmona
Place of Publication	Maspaloma, Spain
Publisher	SPIE
Pages	273-284
Number of pages	12
ISBN (Print)	0-8194-4979-2
DOIs	https://doi.org/10.1117/12.499549
Publication status	Published - May 2003
Event	Bioengineered and Bioinspired Systems - Gran Canaria, Spain Duration: 19 May 2003 → 21 May 2003

Publication series

Name	Proceedings of SPIE
Publisher	SPIE
Volume	5119

Conference

Conference	Bioengineered and Bioinspired Systems
Period	19/05/03 → 21/05/03
Other	19-21 May 2003

Keywords

METIS-211890
Spatial Computing
Wave Computing
Field-Programmable Gate-Array
Feed-Forward Neural Network
EWI-9668
Cellular Neural Network. Temporal computing
Reconfiguration
Modularity
IR-45308

Access to Document

10.1117/12.499549

Cite this

@inproceedings{91016d5a933d427494b41bef1e7b9303,

title = "Natural learning of neural networks by reconfiguration",

abstract = "The communicational and computational demands of neural networks are hard to satisfy in a digital technology. Temporal computing addresses this problem by iteration, but leaves a slow network. Spatial computing only became an option with the coming of modern FPGA devices. The paper provides two examples. First the balance between area and time is discussed on the realization of a modular feed-forward network. Second, the design of real-time image processing through a Cellular Neural Network is treated. In both examples, reconfiguration can be applied to provide for a natural and transparent support of learning.",

keywords = "METIS-211890, Spatial Computing, Wave Computing, Field-Programmable Gate-Array, Feed-Forward Neural Network, EWI-9668, Cellular Neural Network. Temporal computing, Reconfiguration, Modularity, IR-45308",

author = "L. Spaanenburg and R. Alberts and Slump, {Cornelis H.} and {van der Zwaag}, B.J.",

note = "SAS 11-03 ; null ; Conference date: 19-05-2003 Through 21-05-2003",

year = "2003",

month = may,

doi = "10.1117/12.499549",

language = "Undefined",

isbn = "0-8194-4979-2",

series = "Proceedings of SPIE",

publisher = "SPIE",

pages = "273--284",

editor = "Angel Rodriguez-Vazquez and Derek Abbott and Ricardo Carmona",

booktitle = "Bioengineered and Bioinspired Systems",

address = "United States",

}

Natural learning of neural networks by reconfiguration. / Spaanenburg, L.; Alberts, R.; Slump, Cornelis H.; van der Zwaag, B.J.

Bioengineered and Bioinspired Systems. ed. / Angel Rodriguez-Vazquez; Derek Abbott; Ricardo Carmona. Maspaloma, Spain : SPIE, 2003. p. 273-284 (Proceedings of SPIE; Vol. 5119).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Natural learning of neural networks by reconfiguration

AU - Spaanenburg, L.

AU - Alberts, R.

AU - Slump, Cornelis H.

AU - van der Zwaag, B.J.

N1 - SAS 11-03

PY - 2003/5

Y1 - 2003/5

N2 - The communicational and computational demands of neural networks are hard to satisfy in a digital technology. Temporal computing addresses this problem by iteration, but leaves a slow network. Spatial computing only became an option with the coming of modern FPGA devices. The paper provides two examples. First the balance between area and time is discussed on the realization of a modular feed-forward network. Second, the design of real-time image processing through a Cellular Neural Network is treated. In both examples, reconfiguration can be applied to provide for a natural and transparent support of learning.

AB - The communicational and computational demands of neural networks are hard to satisfy in a digital technology. Temporal computing addresses this problem by iteration, but leaves a slow network. Spatial computing only became an option with the coming of modern FPGA devices. The paper provides two examples. First the balance between area and time is discussed on the realization of a modular feed-forward network. Second, the design of real-time image processing through a Cellular Neural Network is treated. In both examples, reconfiguration can be applied to provide for a natural and transparent support of learning.

KW - METIS-211890

KW - Spatial Computing

KW - Wave Computing

KW - Field-Programmable Gate-Array

KW - Feed-Forward Neural Network

KW - EWI-9668

KW - Cellular Neural Network. Temporal computing

KW - Reconfiguration

KW - Modularity

KW - IR-45308

U2 - 10.1117/12.499549

DO - 10.1117/12.499549

M3 - Conference contribution

SN - 0-8194-4979-2

T3 - Proceedings of SPIE

SP - 273

EP - 284

BT - Bioengineered and Bioinspired Systems

A2 - Rodriguez-Vazquez, Angel

A2 - Abbott, Derek

A2 - Carmona, Ricardo

PB - SPIE

CY - Maspaloma, Spain

Y2 - 19 May 2003 through 21 May 2003

ER -