On taxonomies for multi-class image categorization

Alexander Binder, Klaus Muller, Motoaki Kawanabe

Research output: Contribution to journalArticle

30 Citations (Scopus)

Abstract

We study the problem of classifying images into a given, pre-determined taxonomy. This task can be elegantly translated into the structured learning framework. However, despite its power, structured learning has known limits in scalability due to its high memory requirements and slow training process. We propose an efficient approximation of the structured learning approach by an ensemble of local support vector machines (SVMs) that can be trained efficiently with standard techniques. A first theoretical discussion and experiments on toy-data allow to shed light onto why taxonomy-based classification can outperform taxonomy-free approaches and why an appropriately combined ensemble of local SVMs might be of high practical use. Further empirical results on subsets of Caltech256 and VOC2006 data indeed show that our local SVM formulation can effectively exploit the taxonomy structure and thus outperforms standard multi-class classification algorithms while it achieves on par results with taxonomy-based structured algorithms at a significantly decreased computing time.

Original languageEnglish
Pages (from-to)281-301
Number of pages21
JournalInternational Journal of Computer Vision
Volume99
Issue number3
DOIs
Publication statusPublished - 2012 Sep 1
Externally publishedYes

Fingerprint

Taxonomies
Support vector machines
Scalability
Data storage equipment
Experiments

Keywords

  • Multi-class object categorization
  • Structure learning
  • Support vector machine
  • Taxonomies

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence
  • Computer Vision and Pattern Recognition

Cite this

On taxonomies for multi-class image categorization. / Binder, Alexander; Muller, Klaus; Kawanabe, Motoaki.

In: International Journal of Computer Vision, Vol. 99, No. 3, 01.09.2012, p. 281-301.

Research output: Contribution to journalArticle

Binder, Alexander ; Muller, Klaus ; Kawanabe, Motoaki. / On taxonomies for multi-class image categorization. In: International Journal of Computer Vision. 2012 ; Vol. 99, No. 3. pp. 281-301.
@article{90c3cd9a2dee4083abb21605f81404b4,
title = "On taxonomies for multi-class image categorization",
abstract = "We study the problem of classifying images into a given, pre-determined taxonomy. This task can be elegantly translated into the structured learning framework. However, despite its power, structured learning has known limits in scalability due to its high memory requirements and slow training process. We propose an efficient approximation of the structured learning approach by an ensemble of local support vector machines (SVMs) that can be trained efficiently with standard techniques. A first theoretical discussion and experiments on toy-data allow to shed light onto why taxonomy-based classification can outperform taxonomy-free approaches and why an appropriately combined ensemble of local SVMs might be of high practical use. Further empirical results on subsets of Caltech256 and VOC2006 data indeed show that our local SVM formulation can effectively exploit the taxonomy structure and thus outperforms standard multi-class classification algorithms while it achieves on par results with taxonomy-based structured algorithms at a significantly decreased computing time.",
keywords = "Multi-class object categorization, Structure learning, Support vector machine, Taxonomies",
author = "Alexander Binder and Klaus Muller and Motoaki Kawanabe",
year = "2012",
month = "9",
day = "1",
doi = "10.1007/s11263-010-0417-8",
language = "English",
volume = "99",
pages = "281--301",
journal = "International Journal of Computer Vision",
issn = "0920-5691",
publisher = "Springer Netherlands",
number = "3",

}

TY - JOUR

T1 - On taxonomies for multi-class image categorization

AU - Binder, Alexander

AU - Muller, Klaus

AU - Kawanabe, Motoaki

PY - 2012/9/1

Y1 - 2012/9/1

N2 - We study the problem of classifying images into a given, pre-determined taxonomy. This task can be elegantly translated into the structured learning framework. However, despite its power, structured learning has known limits in scalability due to its high memory requirements and slow training process. We propose an efficient approximation of the structured learning approach by an ensemble of local support vector machines (SVMs) that can be trained efficiently with standard techniques. A first theoretical discussion and experiments on toy-data allow to shed light onto why taxonomy-based classification can outperform taxonomy-free approaches and why an appropriately combined ensemble of local SVMs might be of high practical use. Further empirical results on subsets of Caltech256 and VOC2006 data indeed show that our local SVM formulation can effectively exploit the taxonomy structure and thus outperforms standard multi-class classification algorithms while it achieves on par results with taxonomy-based structured algorithms at a significantly decreased computing time.

AB - We study the problem of classifying images into a given, pre-determined taxonomy. This task can be elegantly translated into the structured learning framework. However, despite its power, structured learning has known limits in scalability due to its high memory requirements and slow training process. We propose an efficient approximation of the structured learning approach by an ensemble of local support vector machines (SVMs) that can be trained efficiently with standard techniques. A first theoretical discussion and experiments on toy-data allow to shed light onto why taxonomy-based classification can outperform taxonomy-free approaches and why an appropriately combined ensemble of local SVMs might be of high practical use. Further empirical results on subsets of Caltech256 and VOC2006 data indeed show that our local SVM formulation can effectively exploit the taxonomy structure and thus outperforms standard multi-class classification algorithms while it achieves on par results with taxonomy-based structured algorithms at a significantly decreased computing time.

KW - Multi-class object categorization

KW - Structure learning

KW - Support vector machine

KW - Taxonomies

UR - http://www.scopus.com/inward/record.url?scp=84863619408&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84863619408&partnerID=8YFLogxK

U2 - 10.1007/s11263-010-0417-8

DO - 10.1007/s11263-010-0417-8

M3 - Article

VL - 99

SP - 281

EP - 301

JO - International Journal of Computer Vision

JF - International Journal of Computer Vision

SN - 0920-5691

IS - 3

ER -