Recognition with local features: The kernel recipe

Christian Wallraven, Barbara Caputo, Arnulf Graf

Research output: Chapter in Book/Report/Conference proceedingConference contribution

256 Citations (Scopus)

Abstract

Recent developments in computer vision have shown thai local features can provide efficient representations suitable for robust object recognition. Support Vector Machines have been established as powerful learning algorithms with good generalization capabilities. In this paper, we combine these two approaches and propose a general kernel method for recognition with local features. We show that the proposed kernel satisfies the Mercer condition and that it is suitable for many established local feature frameworks. Large-scale recognition results are presented on three different databases, which demonstrate that SVMs with the proposed kernel perform better than standard matching techniques on local features. In addition, experiments on noisy and occluded images show that local feature representations significantly outperform global approaches.

Original languageEnglish
Title of host publicationProceedings of the IEEE International Conference on Computer Vision
Pages257-264
Number of pages8
Volume1
Publication statusPublished - 2003 Dec 2
Externally publishedYes
EventProceedings: Ninth IEEE International Conference on Computer Vision - Nice, France
Duration: 2003 Oct 132003 Oct 16

Other

OtherProceedings: Ninth IEEE International Conference on Computer Vision
CountryFrance
CityNice
Period03/10/1303/10/16

Fingerprint

Object recognition
Learning algorithms
Computer vision
Support vector machines
Experiments

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition

Cite this

Wallraven, C., Caputo, B., & Graf, A. (2003). Recognition with local features: The kernel recipe. In Proceedings of the IEEE International Conference on Computer Vision (Vol. 1, pp. 257-264)

Recognition with local features : The kernel recipe. / Wallraven, Christian; Caputo, Barbara; Graf, Arnulf.

Proceedings of the IEEE International Conference on Computer Vision. Vol. 1 2003. p. 257-264.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Wallraven, C, Caputo, B & Graf, A 2003, Recognition with local features: The kernel recipe. in Proceedings of the IEEE International Conference on Computer Vision. vol. 1, pp. 257-264, Proceedings: Ninth IEEE International Conference on Computer Vision, Nice, France, 03/10/13.
Wallraven C, Caputo B, Graf A. Recognition with local features: The kernel recipe. In Proceedings of the IEEE International Conference on Computer Vision. Vol. 1. 2003. p. 257-264
Wallraven, Christian ; Caputo, Barbara ; Graf, Arnulf. / Recognition with local features : The kernel recipe. Proceedings of the IEEE International Conference on Computer Vision. Vol. 1 2003. pp. 257-264
@inproceedings{bc1fbaa351ea451ba436a0544ef3a60a,
title = "Recognition with local features: The kernel recipe",
abstract = "Recent developments in computer vision have shown thai local features can provide efficient representations suitable for robust object recognition. Support Vector Machines have been established as powerful learning algorithms with good generalization capabilities. In this paper, we combine these two approaches and propose a general kernel method for recognition with local features. We show that the proposed kernel satisfies the Mercer condition and that it is suitable for many established local feature frameworks. Large-scale recognition results are presented on three different databases, which demonstrate that SVMs with the proposed kernel perform better than standard matching techniques on local features. In addition, experiments on noisy and occluded images show that local feature representations significantly outperform global approaches.",
author = "Christian Wallraven and Barbara Caputo and Arnulf Graf",
year = "2003",
month = "12",
day = "2",
language = "English",
volume = "1",
pages = "257--264",
booktitle = "Proceedings of the IEEE International Conference on Computer Vision",

}

TY - GEN

T1 - Recognition with local features

T2 - The kernel recipe

AU - Wallraven, Christian

AU - Caputo, Barbara

AU - Graf, Arnulf

PY - 2003/12/2

Y1 - 2003/12/2

N2 - Recent developments in computer vision have shown thai local features can provide efficient representations suitable for robust object recognition. Support Vector Machines have been established as powerful learning algorithms with good generalization capabilities. In this paper, we combine these two approaches and propose a general kernel method for recognition with local features. We show that the proposed kernel satisfies the Mercer condition and that it is suitable for many established local feature frameworks. Large-scale recognition results are presented on three different databases, which demonstrate that SVMs with the proposed kernel perform better than standard matching techniques on local features. In addition, experiments on noisy and occluded images show that local feature representations significantly outperform global approaches.

AB - Recent developments in computer vision have shown thai local features can provide efficient representations suitable for robust object recognition. Support Vector Machines have been established as powerful learning algorithms with good generalization capabilities. In this paper, we combine these two approaches and propose a general kernel method for recognition with local features. We show that the proposed kernel satisfies the Mercer condition and that it is suitable for many established local feature frameworks. Large-scale recognition results are presented on three different databases, which demonstrate that SVMs with the proposed kernel perform better than standard matching techniques on local features. In addition, experiments on noisy and occluded images show that local feature representations significantly outperform global approaches.

UR - http://www.scopus.com/inward/record.url?scp=0345414121&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0345414121&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0345414121

VL - 1

SP - 257

EP - 264

BT - Proceedings of the IEEE International Conference on Computer Vision

ER -