Improving image classification by co-training with multi-modal features

We explore the use of co-training to improve the performance of image classification in the setting where multiple classifiers are used and several types of features are available. Features are assigned to classifiers in an optimal manner using hierarchical clustering with a distance metric based on...

Full description

Bibliographic Details
Main Author: Weston, Kyle
Other Authors: Martin D Levine (Internal/Supervisor)
Format: Others
Language:en
Published: McGill University 2011
Subjects:
Online Access:http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=104797
id ndltd-LACETR-oai-collectionscanada.gc.ca-QMM.104797
record_format oai_dc
spelling ndltd-LACETR-oai-collectionscanada.gc.ca-QMM.1047972014-02-13T03:48:52ZImproving image classification by co-training with multi-modal featuresWeston, KyleApplied Sciences - Artificial IntelligenceWe explore the use of co-training to improve the performance of image classification in the setting where multiple classifiers are used and several types of features are available. Features are assigned to classifiers in an optimal manner using hierarchical clustering with a distance metric based on conditional mutual information. The effect of increasing the number of classifiers is then evaluated by co-training using the assigned feature sets. Experimental results indicate that the feature assignments chosen by the clustering approach afford superior co-training performance in comparison to other logical assignment choices. The results also indicate that increasing the number of classifiers beyond two leads to improved performance provided that the classifiers are sufficiently independent, and are reasonable well balanced in terms of labeling ability.Additionally, we explore the effect that the initial training set selectionhas on co-training performance. We find that the quality of training imageshas a profound effect on performance and provide recommendations for howbest to select these images.Nous explorons l'utilisation de la co-formation pour améliorer la performance de classification d'image dans un milieu où multiples classificateurs s'emploient et plusieurs types de caractéristiques sont disponibles. Les caractéristiques sont associés aux classificateurs d'une manière optimal en employant le groupage hiérarchique avec une mesure de distance basée sur l'information mutuelle conditionnelle. L'effet d'augmenter le nombre de classificateurs est alors evalué par la co-formation, en employant les ensembles de caractéristiques attribués. Les résultats de nos expériences indique que si on augmente le nombre de classificateurs au-delà de deux, la performance s'améliore pourvu que les caractéristiques soient suffisamment indépendantes et assez bien équilibrées en termes de compétence d'étiquetage. En plus, nous explorons l'effet de l'ensemble choisi pour l'entraînement initial sur la performance en co-formation. Nous trouvons que la qualité d'images dans l'entraînement a un effet profond sur la performance, et nous fournissons des recommandations sur comment sélectionner ces images pour le meilleur effet.McGill UniversityMartin D Levine (Internal/Supervisor)2011Electronic Thesis or Dissertationapplication/pdfenElectronically-submitted theses.All items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated.Master of Engineering (Department of Electrical and Computer Engineering) http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=104797
collection NDLTD
language en
format Others
sources NDLTD
topic Applied Sciences - Artificial Intelligence
spellingShingle Applied Sciences - Artificial Intelligence
Weston, Kyle
Improving image classification by co-training with multi-modal features
description We explore the use of co-training to improve the performance of image classification in the setting where multiple classifiers are used and several types of features are available. Features are assigned to classifiers in an optimal manner using hierarchical clustering with a distance metric based on conditional mutual information. The effect of increasing the number of classifiers is then evaluated by co-training using the assigned feature sets. Experimental results indicate that the feature assignments chosen by the clustering approach afford superior co-training performance in comparison to other logical assignment choices. The results also indicate that increasing the number of classifiers beyond two leads to improved performance provided that the classifiers are sufficiently independent, and are reasonable well balanced in terms of labeling ability.Additionally, we explore the effect that the initial training set selectionhas on co-training performance. We find that the quality of training imageshas a profound effect on performance and provide recommendations for howbest to select these images. === Nous explorons l'utilisation de la co-formation pour améliorer la performance de classification d'image dans un milieu où multiples classificateurs s'emploient et plusieurs types de caractéristiques sont disponibles. Les caractéristiques sont associés aux classificateurs d'une manière optimal en employant le groupage hiérarchique avec une mesure de distance basée sur l'information mutuelle conditionnelle. L'effet d'augmenter le nombre de classificateurs est alors evalué par la co-formation, en employant les ensembles de caractéristiques attribués. Les résultats de nos expériences indique que si on augmente le nombre de classificateurs au-delà de deux, la performance s'améliore pourvu que les caractéristiques soient suffisamment indépendantes et assez bien équilibrées en termes de compétence d'étiquetage. En plus, nous explorons l'effet de l'ensemble choisi pour l'entraînement initial sur la performance en co-formation. Nous trouvons que la qualité d'images dans l'entraînement a un effet profond sur la performance, et nous fournissons des recommandations sur comment sélectionner ces images pour le meilleur effet.
author2 Martin D Levine (Internal/Supervisor)
author_facet Martin D Levine (Internal/Supervisor)
Weston, Kyle
author Weston, Kyle
author_sort Weston, Kyle
title Improving image classification by co-training with multi-modal features
title_short Improving image classification by co-training with multi-modal features
title_full Improving image classification by co-training with multi-modal features
title_fullStr Improving image classification by co-training with multi-modal features
title_full_unstemmed Improving image classification by co-training with multi-modal features
title_sort improving image classification by co-training with multi-modal features
publisher McGill University
publishDate 2011
url http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=104797
work_keys_str_mv AT westonkyle improvingimageclassificationbycotrainingwithmultimodalfeatures
_version_ 1716639479664476160