Stacking of texture based filters for visual place categorization

Nur Nabilah Abu Mangshor, Azizi Abdullah

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

Abstract

Recent research in computer vision has shown that combining multiple features is an effective way to improve classification performance. Furthermore, the use of filters that convolve images at multiple filter responses can increase descriptions to images. The more distinctive the filter responses, the better it able to distinguish characteristics from other groups. Thus, this paper describes a combination method that combines multiple classifier outputs at several filter responses to enhance the automatic visual place categorization system. Besides, one of the goals of this study is to explore performance differences between single and dedicated combination of filter response classifier methods. One possible problem of combining multiple filter responses for describing images is that the input vector becomes very large in dimensionality, which can increase the problem of overfitting and hinder generalization performance. Therefore, the stacking of support vector machine is used to compute the right output class from each single descriptor of filter responses that has been trained at the first layer of support vector machine. Next, the second layer support vector machine is used to combine those class probability output values of all trained first layer support vector models to learn the right output class. We have performed experiments on five different categories of visual places from the KTH-IDOL2 dataset with a single descriptor using 25 different filter responses of Laws filters. Results showed that the 2-layer stacking algorithm outperform the single and naive approaches that uses single filter response input vector and combines all filter response outputs directly in a very large single input vector respectively.

Original languageEnglish
Title of host publication2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages354-359
Number of pages6
ISBN (Print)9781479934003
DOIs
Publication statusPublished - 3 Mar 2003
Event2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013 - Hanoi, Viet Nam
Duration: 15 Dec 201318 Dec 2013

Other

Other2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013
CountryViet Nam
CityHanoi
Period15/12/1318/12/13

Fingerprint

Textures
Support vector machines
Classifiers
Computer vision
Experiments

Keywords

  • edge histogram descriptor
  • Laws filters
  • naive approach
  • stacking approach
  • support vector machine
  • Visual place categorization

ASJC Scopus subject areas

  • Computer Science Applications
  • Computer Vision and Pattern Recognition
  • Software

Cite this

Mangshor, N. N. A., & Abdullah, A. (2003). Stacking of texture based filters for visual place categorization. In 2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013 (pp. 354-359). [7054158] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SOCPAR.2013.7054158

Stacking of texture based filters for visual place categorization. / Mangshor, Nur Nabilah Abu; Abdullah, Azizi.

2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013. Institute of Electrical and Electronics Engineers Inc., 2003. p. 354-359 7054158.

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

Mangshor, NNA & Abdullah, A 2003, Stacking of texture based filters for visual place categorization. in 2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013., 7054158, Institute of Electrical and Electronics Engineers Inc., pp. 354-359, 2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013, Hanoi, Viet Nam, 15/12/13. https://doi.org/10.1109/SOCPAR.2013.7054158
Mangshor NNA, Abdullah A. Stacking of texture based filters for visual place categorization. In 2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013. Institute of Electrical and Electronics Engineers Inc. 2003. p. 354-359. 7054158 https://doi.org/10.1109/SOCPAR.2013.7054158
Mangshor, Nur Nabilah Abu ; Abdullah, Azizi. / Stacking of texture based filters for visual place categorization. 2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013. Institute of Electrical and Electronics Engineers Inc., 2003. pp. 354-359
@inproceedings{d87bd6aad28f41af988da7a53177f548,
title = "Stacking of texture based filters for visual place categorization",
abstract = "Recent research in computer vision has shown that combining multiple features is an effective way to improve classification performance. Furthermore, the use of filters that convolve images at multiple filter responses can increase descriptions to images. The more distinctive the filter responses, the better it able to distinguish characteristics from other groups. Thus, this paper describes a combination method that combines multiple classifier outputs at several filter responses to enhance the automatic visual place categorization system. Besides, one of the goals of this study is to explore performance differences between single and dedicated combination of filter response classifier methods. One possible problem of combining multiple filter responses for describing images is that the input vector becomes very large in dimensionality, which can increase the problem of overfitting and hinder generalization performance. Therefore, the stacking of support vector machine is used to compute the right output class from each single descriptor of filter responses that has been trained at the first layer of support vector machine. Next, the second layer support vector machine is used to combine those class probability output values of all trained first layer support vector models to learn the right output class. We have performed experiments on five different categories of visual places from the KTH-IDOL2 dataset with a single descriptor using 25 different filter responses of Laws filters. Results showed that the 2-layer stacking algorithm outperform the single and naive approaches that uses single filter response input vector and combines all filter response outputs directly in a very large single input vector respectively.",
keywords = "edge histogram descriptor, Laws filters, naive approach, stacking approach, support vector machine, Visual place categorization",
author = "Mangshor, {Nur Nabilah Abu} and Azizi Abdullah",
year = "2003",
month = "3",
day = "3",
doi = "10.1109/SOCPAR.2013.7054158",
language = "English",
isbn = "9781479934003",
pages = "354--359",
booktitle = "2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Stacking of texture based filters for visual place categorization

AU - Mangshor, Nur Nabilah Abu

AU - Abdullah, Azizi

PY - 2003/3/3

Y1 - 2003/3/3

N2 - Recent research in computer vision has shown that combining multiple features is an effective way to improve classification performance. Furthermore, the use of filters that convolve images at multiple filter responses can increase descriptions to images. The more distinctive the filter responses, the better it able to distinguish characteristics from other groups. Thus, this paper describes a combination method that combines multiple classifier outputs at several filter responses to enhance the automatic visual place categorization system. Besides, one of the goals of this study is to explore performance differences between single and dedicated combination of filter response classifier methods. One possible problem of combining multiple filter responses for describing images is that the input vector becomes very large in dimensionality, which can increase the problem of overfitting and hinder generalization performance. Therefore, the stacking of support vector machine is used to compute the right output class from each single descriptor of filter responses that has been trained at the first layer of support vector machine. Next, the second layer support vector machine is used to combine those class probability output values of all trained first layer support vector models to learn the right output class. We have performed experiments on five different categories of visual places from the KTH-IDOL2 dataset with a single descriptor using 25 different filter responses of Laws filters. Results showed that the 2-layer stacking algorithm outperform the single and naive approaches that uses single filter response input vector and combines all filter response outputs directly in a very large single input vector respectively.

AB - Recent research in computer vision has shown that combining multiple features is an effective way to improve classification performance. Furthermore, the use of filters that convolve images at multiple filter responses can increase descriptions to images. The more distinctive the filter responses, the better it able to distinguish characteristics from other groups. Thus, this paper describes a combination method that combines multiple classifier outputs at several filter responses to enhance the automatic visual place categorization system. Besides, one of the goals of this study is to explore performance differences between single and dedicated combination of filter response classifier methods. One possible problem of combining multiple filter responses for describing images is that the input vector becomes very large in dimensionality, which can increase the problem of overfitting and hinder generalization performance. Therefore, the stacking of support vector machine is used to compute the right output class from each single descriptor of filter responses that has been trained at the first layer of support vector machine. Next, the second layer support vector machine is used to combine those class probability output values of all trained first layer support vector models to learn the right output class. We have performed experiments on five different categories of visual places from the KTH-IDOL2 dataset with a single descriptor using 25 different filter responses of Laws filters. Results showed that the 2-layer stacking algorithm outperform the single and naive approaches that uses single filter response input vector and combines all filter response outputs directly in a very large single input vector respectively.

KW - edge histogram descriptor

KW - Laws filters

KW - naive approach

KW - stacking approach

KW - support vector machine

KW - Visual place categorization

UR - http://www.scopus.com/inward/record.url?scp=84949926144&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84949926144&partnerID=8YFLogxK

U2 - 10.1109/SOCPAR.2013.7054158

DO - 10.1109/SOCPAR.2013.7054158

M3 - Conference contribution

AN - SCOPUS:84949926144

SN - 9781479934003

SP - 354

EP - 359

BT - 2013 International Conference on Soft Computing and Pattern Recognition, SoCPaR 2013

PB - Institute of Electrical and Electronics Engineers Inc.

ER -