On the analysis of neural networks for image processing

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

45 Downloads (Pure)

Abstract

This paper illustrates a novel method to analyze artificial neural networks so as to gain insight into their internal functionality. To this purpose, we will show analysis results of some feed-forward–error-back-propagation neural networks for image processing. We will describe them in terms of domain-dependent basic functions, which are, in the case of the digital image processing domain, differential operators of various orders and with various angles of operation. Some other pixel classification techniques are analyzed in the same way, enabling easy comparison.
Original languageUndefined
Title of host publicationProceedings of KES2003 vol. II
EditorsVasile Palade, Robert J. Howlett, Lakhmi Jain
Place of PublicationOxford, UK
PublisherSpringer
Pages950-957
Number of pages8
ISBN (Print)3-540-40804-5
DOIs
Publication statusPublished - Sep 2003

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume2774

Keywords

  • IR-45400
  • METIS-212150
  • EWI-9667

Cite this

van der Zwaag, B. J., Slump, C. H., & Spaanenburg, L. (2003). On the analysis of neural networks for image processing. In V. Palade, R. J. Howlett, & L. Jain (Eds.), Proceedings of KES2003 vol. II (pp. 950-957). (Lecture Notes in Computer Science; Vol. 2774). Oxford, UK: Springer. https://doi.org/10.1007/978-3-540-45226-3_130, https://doi.org/10.1007/b12003
van der Zwaag, B.J. ; Slump, Cornelis H. ; Spaanenburg, Lambert. / On the analysis of neural networks for image processing. Proceedings of KES2003 vol. II. editor / Vasile Palade ; Robert J. Howlett ; Lakhmi Jain. Oxford, UK : Springer, 2003. pp. 950-957 (Lecture Notes in Computer Science).
@inproceedings{b1edccba19f44f27a967d9654a70f6a8,
title = "On the analysis of neural networks for image processing",
abstract = "This paper illustrates a novel method to analyze artificial neural networks so as to gain insight into their internal functionality. To this purpose, we will show analysis results of some feed-forward–error-back-propagation neural networks for image processing. We will describe them in terms of domain-dependent basic functions, which are, in the case of the digital image processing domain, differential operators of various orders and with various angles of operation. Some other pixel classification techniques are analyzed in the same way, enabling easy comparison.",
keywords = "IR-45400, METIS-212150, EWI-9667",
author = "{van der Zwaag}, B.J. and Slump, {Cornelis H.} and Lambert Spaanenburg",
note = "SAS 27-03",
year = "2003",
month = "9",
doi = "10.1007/978-3-540-45226-3_130",
language = "Undefined",
isbn = "3-540-40804-5",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "950--957",
editor = "Vasile Palade and Howlett, {Robert J.} and Lakhmi Jain",
booktitle = "Proceedings of KES2003 vol. II",

}

van der Zwaag, BJ, Slump, CH & Spaanenburg, L 2003, On the analysis of neural networks for image processing. in V Palade, RJ Howlett & L Jain (eds), Proceedings of KES2003 vol. II. Lecture Notes in Computer Science, vol. 2774, Springer, Oxford, UK, pp. 950-957. https://doi.org/10.1007/978-3-540-45226-3_130, https://doi.org/10.1007/b12003

On the analysis of neural networks for image processing. / van der Zwaag, B.J.; Slump, Cornelis H.; Spaanenburg, Lambert.

Proceedings of KES2003 vol. II. ed. / Vasile Palade; Robert J. Howlett; Lakhmi Jain. Oxford, UK : Springer, 2003. p. 950-957 (Lecture Notes in Computer Science; Vol. 2774).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - On the analysis of neural networks for image processing

AU - van der Zwaag, B.J.

AU - Slump, Cornelis H.

AU - Spaanenburg, Lambert

N1 - SAS 27-03

PY - 2003/9

Y1 - 2003/9

N2 - This paper illustrates a novel method to analyze artificial neural networks so as to gain insight into their internal functionality. To this purpose, we will show analysis results of some feed-forward–error-back-propagation neural networks for image processing. We will describe them in terms of domain-dependent basic functions, which are, in the case of the digital image processing domain, differential operators of various orders and with various angles of operation. Some other pixel classification techniques are analyzed in the same way, enabling easy comparison.

AB - This paper illustrates a novel method to analyze artificial neural networks so as to gain insight into their internal functionality. To this purpose, we will show analysis results of some feed-forward–error-back-propagation neural networks for image processing. We will describe them in terms of domain-dependent basic functions, which are, in the case of the digital image processing domain, differential operators of various orders and with various angles of operation. Some other pixel classification techniques are analyzed in the same way, enabling easy comparison.

KW - IR-45400

KW - METIS-212150

KW - EWI-9667

U2 - 10.1007/978-3-540-45226-3_130

DO - 10.1007/978-3-540-45226-3_130

M3 - Conference contribution

SN - 3-540-40804-5

T3 - Lecture Notes in Computer Science

SP - 950

EP - 957

BT - Proceedings of KES2003 vol. II

A2 - Palade, Vasile

A2 - Howlett, Robert J.

A2 - Jain, Lakhmi

PB - Springer

CY - Oxford, UK

ER -

van der Zwaag BJ, Slump CH, Spaanenburg L. On the analysis of neural networks for image processing. In Palade V, Howlett RJ, Jain L, editors, Proceedings of KES2003 vol. II. Oxford, UK: Springer. 2003. p. 950-957. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-540-45226-3_130, https://doi.org/10.1007/b12003