Tampere University of Technology

TUTCRIS Research Portal

An ensemble of classifiers based on different texture descriptors for texture classification

Research output: Contribution to journalArticleScientificpeer-review

Standard

An ensemble of classifiers based on different texture descriptors for texture classification. / Paci, Michelangelo; Nanni, Loris; Severi, Stefano.

In: Journal of King Saud University - Science, Vol. 25, No. 3, 07.2013, p. 235-244.

Research output: Contribution to journalArticleScientificpeer-review

Harvard

Paci, M, Nanni, L & Severi, S 2013, 'An ensemble of classifiers based on different texture descriptors for texture classification', Journal of King Saud University - Science, vol. 25, no. 3, pp. 235-244. https://doi.org/10.1016/j.jksus.2012.12.001

APA

Vancouver

Author

Paci, Michelangelo ; Nanni, Loris ; Severi, Stefano. / An ensemble of classifiers based on different texture descriptors for texture classification. In: Journal of King Saud University - Science. 2013 ; Vol. 25, No. 3. pp. 235-244.

Bibtex - Download

@article{e21a1a197bb04418854fa4a608585bd9,
title = "An ensemble of classifiers based on different texture descriptors for texture classification",
abstract = "Here we propose a system that incorporates two different state-of-the-art classifiers (support vector machine and gaussian process classifier) and two different descriptors (multi local quinary patterns and multi local phase quantization with ternary coding) for texture classification.Both the tested descriptors are an ensemble of stand-alone descriptors obtained using different parameters setting (the same set is used in each dataset). For each stand-alone descriptor we train a different classifier, the set of scores of each classifier is normalized to mean equal to zero and standard deviation equal to one, then all the score sets are combined by the sum rule.Our experimental section shows that we succeed in building a high performance ensemble that works well on different datasets without any ad hoc parameters tuning. The fusion among the different systems permits to outperform SVM where the parameters and kernels are tuned separately in each dataset, while in the proposed ensemble the linear SVM, with the same parameter cost in all the datasets, is used.",
keywords = "Ensemble of classifiers, Machine learning, Non-binary coding, Support vector machine, Texture descriptors",
author = "Michelangelo Paci and Loris Nanni and Stefano Severi",
year = "2013",
month = "7",
doi = "10.1016/j.jksus.2012.12.001",
language = "English",
volume = "25",
pages = "235--244",
journal = "Journal of King Saud University - Science",
issn = "1018-3647",
publisher = "King Saud University",
number = "3",

}

RIS (suitable for import to EndNote) - Download

TY - JOUR

T1 - An ensemble of classifiers based on different texture descriptors for texture classification

AU - Paci, Michelangelo

AU - Nanni, Loris

AU - Severi, Stefano

PY - 2013/7

Y1 - 2013/7

N2 - Here we propose a system that incorporates two different state-of-the-art classifiers (support vector machine and gaussian process classifier) and two different descriptors (multi local quinary patterns and multi local phase quantization with ternary coding) for texture classification.Both the tested descriptors are an ensemble of stand-alone descriptors obtained using different parameters setting (the same set is used in each dataset). For each stand-alone descriptor we train a different classifier, the set of scores of each classifier is normalized to mean equal to zero and standard deviation equal to one, then all the score sets are combined by the sum rule.Our experimental section shows that we succeed in building a high performance ensemble that works well on different datasets without any ad hoc parameters tuning. The fusion among the different systems permits to outperform SVM where the parameters and kernels are tuned separately in each dataset, while in the proposed ensemble the linear SVM, with the same parameter cost in all the datasets, is used.

AB - Here we propose a system that incorporates two different state-of-the-art classifiers (support vector machine and gaussian process classifier) and two different descriptors (multi local quinary patterns and multi local phase quantization with ternary coding) for texture classification.Both the tested descriptors are an ensemble of stand-alone descriptors obtained using different parameters setting (the same set is used in each dataset). For each stand-alone descriptor we train a different classifier, the set of scores of each classifier is normalized to mean equal to zero and standard deviation equal to one, then all the score sets are combined by the sum rule.Our experimental section shows that we succeed in building a high performance ensemble that works well on different datasets without any ad hoc parameters tuning. The fusion among the different systems permits to outperform SVM where the parameters and kernels are tuned separately in each dataset, while in the proposed ensemble the linear SVM, with the same parameter cost in all the datasets, is used.

KW - Ensemble of classifiers

KW - Machine learning

KW - Non-binary coding

KW - Support vector machine

KW - Texture descriptors

UR - http://www.scopus.com/inward/record.url?scp=84878964801&partnerID=8YFLogxK

U2 - 10.1016/j.jksus.2012.12.001

DO - 10.1016/j.jksus.2012.12.001

M3 - Article

VL - 25

SP - 235

EP - 244

JO - Journal of King Saud University - Science

JF - Journal of King Saud University - Science

SN - 1018-3647

IS - 3

ER -