Deterministic defuzzification based on Spectral Projected Gradient optimization

Joakim Lindblad; Natasa Sladoje; Tibor Lukic

Deterministic defuzzification based on Spectral Projected Gradient optimization

Joakim Lindblad
, Natasa Sladoje
, Tibor Lukic

Publikation: Kapitel i bok/rapport/konferenshandling › Konferensartikel i proceedings › Peer review

Sammanfattning

We apply deterministic optimization based on Spectral Projected Gradient method in combination with concave regularization to solve the minimization problem imposed by defuzzification by feature distance minimization. We compare the performance of the proposed algorithm with the methods previously recommended for the same task, (non-deterministic) simulated annealing and (deterministic) DC based algorithm. The evaluation, including numerical tests performed on synthetic and real images, shows advantages of the new method in terms of speed and flexibility regarding inclusion of additional features in defuzzification. Its relatively low memory requirements allow the application of the suggested method for defuzzification of 3D objects.

Originalspråk	Engelska
Titel på värdpublikation	Pattern Recognition: 30th DAGM Symposium Munich, Germany, June 10-13, 2008 Proceedings
Förlag	SPRINGER-VERLAG BERLIN
Sidor	476-+
Antal sidor	2
ISBN (tryckt)	978-3-540-69320-8
Status	Publicerad - 2008

Publikationsserier

Serie	Lecture Notes in Computer Science
Volym	5096
ISSN	0302-9743

Andra filer och länkar

Permanent länk till denna sida

Citera det här

@inproceedings{5608a06fd2804ea5b900dd4857b76ad1,

title = "Deterministic defuzzification based on Spectral Projected Gradient optimization",

abstract = "We apply deterministic optimization based on Spectral Projected Gradient method in combination with concave regularization to solve the minimization problem imposed by defuzzification by feature distance minimization. We compare the performance of the proposed algorithm with the methods previously recommended for the same task, (non-deterministic) simulated annealing and (deterministic) DC based algorithm. The evaluation, including numerical tests performed on synthetic and real images, shows advantages of the new method in terms of speed and flexibility regarding inclusion of additional features in defuzzification. Its relatively low memory requirements allow the application of the suggested method for defuzzification of 3D objects.",

author = "Joakim Lindblad and Natasa Sladoje and Tibor Lukic",

year = "2008",

language = "English",

isbn = "978-3-540-69320-8",

series = "Lecture Notes in Computer Science",

publisher = "SPRINGER-VERLAG BERLIN",

pages = "476--+",

booktitle = "Pattern Recognition: 30th DAGM Symposium Munich, Germany, June 10-13, 2008 Proceedings",

}

Deterministic defuzzification based on Spectral Projected Gradient optimization. / Lindblad, Joakim; Sladoje, Natasa; Lukic, Tibor.
Pattern Recognition: 30th DAGM Symposium Munich, Germany, June 10-13, 2008 Proceedings. SPRINGER-VERLAG BERLIN, 2008. s. 476-+ (Lecture Notes in Computer Science, Vol. 5096).

Publikation: Kapitel i bok/rapport/konferenshandling › Konferensartikel i proceedings › Peer review

TY - GEN

T1 - Deterministic defuzzification based on Spectral Projected Gradient optimization

AU - Lindblad, Joakim

AU - Sladoje, Natasa

AU - Lukic, Tibor

PY - 2008

Y1 - 2008

N2 - We apply deterministic optimization based on Spectral Projected Gradient method in combination with concave regularization to solve the minimization problem imposed by defuzzification by feature distance minimization. We compare the performance of the proposed algorithm with the methods previously recommended for the same task, (non-deterministic) simulated annealing and (deterministic) DC based algorithm. The evaluation, including numerical tests performed on synthetic and real images, shows advantages of the new method in terms of speed and flexibility regarding inclusion of additional features in defuzzification. Its relatively low memory requirements allow the application of the suggested method for defuzzification of 3D objects.

AB - We apply deterministic optimization based on Spectral Projected Gradient method in combination with concave regularization to solve the minimization problem imposed by defuzzification by feature distance minimization. We compare the performance of the proposed algorithm with the methods previously recommended for the same task, (non-deterministic) simulated annealing and (deterministic) DC based algorithm. The evaluation, including numerical tests performed on synthetic and real images, shows advantages of the new method in terms of speed and flexibility regarding inclusion of additional features in defuzzification. Its relatively low memory requirements allow the application of the suggested method for defuzzification of 3D objects.

UR - https://res.slu.se/id/publ/21222

M3 - Conference paper in proceedings

SN - 978-3-540-69320-8

T3 - Lecture Notes in Computer Science

SP - 476-+

BT - Pattern Recognition: 30th DAGM Symposium Munich, Germany, June 10-13, 2008 Proceedings

PB - SPRINGER-VERLAG BERLIN

ER -