The GCS kernel for SVM-based image recognition

Sabri Boughorbel; Jean Philippe Tarel; François Fleuret; Nozha Boujemaa

doi:10.1007/11550907_94

The GCS kernel for SVM-based image recognition

Sabri Boughorbel^*
, Jean Philippe Tarel
, François Fleuret
, Nozha Boujemaa

^*Corresponding author for this work

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

11 Citations (Scopus)

Abstract

In this paper, we present a new compactly supported kernel for SVM based image recognition. This kernel which we called Geometric Compactly Supported (GCS) can be viewed as a generalization of spherical kernels to higher dimensions. The construction of the GCS kernel is based on a geometric approach using the intersection volume of two n-dimensional balls. The compactness property of the GCS kernel leads to a sparse Gram matrix which enhances computation efficiency by using sparse linear algebra algorithms. Comparisons of the GCS kernel performance, for image recognition task, with other known kernels prove the interest of this new kernel.

Original language	English
Title of host publication	Artificial Neural Networks
Subtitle of host publication	Formal Models and Their Applications - ICANN 2005 - 15th International Conference Proceedings
Publisher	Springer Verlag
Pages	595-600
Number of pages	6
ISBN (Print)	3540287558, 9783540287551
DOIs	https://doi.org/10.1007/11550907_94
Publication status	Published - 2005
Externally published	Yes
Event	15th International Conference on Artificial Neural Networks: Formal Models and Their Applications, ICANN 2005 - Warsaw, Poland Duration: 11 Sept 2005 → 15 Sept 2005

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3697 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	15th International Conference on Artificial Neural Networks: Formal Models and Their Applications, ICANN 2005
Country/Territory	Poland
City	Warsaw
Period	11/09/05 → 15/09/05

Access to Document

10.1007/11550907_94

Cite this

Boughorbel, S., Tarel, J. P., Fleuret, F., & Boujemaa, N. (2005). The GCS kernel for SVM-based image recognition. In Artificial Neural Networks: Formal Models and Their Applications - ICANN 2005 - 15th International Conference Proceedings (pp. 595-600). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3697 LNCS). Springer Verlag. https://doi.org/10.1007/11550907_94

Boughorbel, Sabri ; Tarel, Jean Philippe ; Fleuret, François et al. / The GCS kernel for SVM-based image recognition. Artificial Neural Networks: Formal Models and Their Applications - ICANN 2005 - 15th International Conference Proceedings. Springer Verlag, 2005. pp. 595-600 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{6aaf4a52d3344d8eade8d8a4cdf7da85,

title = "The GCS kernel for SVM-based image recognition",

abstract = "In this paper, we present a new compactly supported kernel for SVM based image recognition. This kernel which we called Geometric Compactly Supported (GCS) can be viewed as a generalization of spherical kernels to higher dimensions. The construction of the GCS kernel is based on a geometric approach using the intersection volume of two n-dimensional balls. The compactness property of the GCS kernel leads to a sparse Gram matrix which enhances computation efficiency by using sparse linear algebra algorithms. Comparisons of the GCS kernel performance, for image recognition task, with other known kernels prove the interest of this new kernel.",

author = "Sabri Boughorbel and Tarel, \{Jean Philippe\} and Fran{\c c}ois Fleuret and Nozha Boujemaa",

year = "2005",

doi = "10.1007/11550907\_94",

language = "English",

isbn = "3540287558",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "595--600",

booktitle = "Artificial Neural Networks",

address = "Germany",

note = "15th International Conference on Artificial Neural Networks: Formal Models and Their Applications, ICANN 2005 ; Conference date: 11-09-2005 Through 15-09-2005",

}

Boughorbel, S, Tarel, JP, Fleuret, F & Boujemaa, N 2005, The GCS kernel for SVM-based image recognition. in Artificial Neural Networks: Formal Models and Their Applications - ICANN 2005 - 15th International Conference Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3697 LNCS, Springer Verlag, pp. 595-600, 15th International Conference on Artificial Neural Networks: Formal Models and Their Applications, ICANN 2005, Warsaw, Poland, 11/09/05. https://doi.org/10.1007/11550907_94

The GCS kernel for SVM-based image recognition. / Boughorbel, Sabri; Tarel, Jean Philippe; Fleuret, François et al.
Artificial Neural Networks: Formal Models and Their Applications - ICANN 2005 - 15th International Conference Proceedings. Springer Verlag, 2005. p. 595-600 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3697 LNCS).

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

TY - GEN

T1 - The GCS kernel for SVM-based image recognition

AU - Boughorbel, Sabri

AU - Tarel, Jean Philippe

AU - Fleuret, François

AU - Boujemaa, Nozha

PY - 2005

Y1 - 2005

N2 - In this paper, we present a new compactly supported kernel for SVM based image recognition. This kernel which we called Geometric Compactly Supported (GCS) can be viewed as a generalization of spherical kernels to higher dimensions. The construction of the GCS kernel is based on a geometric approach using the intersection volume of two n-dimensional balls. The compactness property of the GCS kernel leads to a sparse Gram matrix which enhances computation efficiency by using sparse linear algebra algorithms. Comparisons of the GCS kernel performance, for image recognition task, with other known kernels prove the interest of this new kernel.

AB - In this paper, we present a new compactly supported kernel for SVM based image recognition. This kernel which we called Geometric Compactly Supported (GCS) can be viewed as a generalization of spherical kernels to higher dimensions. The construction of the GCS kernel is based on a geometric approach using the intersection volume of two n-dimensional balls. The compactness property of the GCS kernel leads to a sparse Gram matrix which enhances computation efficiency by using sparse linear algebra algorithms. Comparisons of the GCS kernel performance, for image recognition task, with other known kernels prove the interest of this new kernel.

UR - https://www.scopus.com/pages/publications/33646255226

U2 - 10.1007/11550907_94

DO - 10.1007/11550907_94

M3 - Conference contribution

AN - SCOPUS:33646255226

SN - 3540287558

SN - 9783540287551

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 595

EP - 600

BT - Artificial Neural Networks

PB - Springer Verlag

T2 - 15th International Conference on Artificial Neural Networks: Formal Models and Their Applications, ICANN 2005

Y2 - 11 September 2005 through 15 September 2005

ER -

Boughorbel S, Tarel JP, Fleuret F, Boujemaa N. The GCS kernel for SVM-based image recognition. In Artificial Neural Networks: Formal Models and Their Applications - ICANN 2005 - 15th International Conference Proceedings. Springer Verlag. 2005. p. 595-600. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/11550907_94

The GCS kernel for SVM-based image recognition

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this