Learning discriminative local features from image-level labelled data for colonoscopy image classification

Siyamalan Manivannan; Emanuele Trucco

doi:10.1109/ISBI.2015.7163901

Learning discriminative local features from image-level labelled data for colonoscopy image classification

Siyamalan Manivannan, Emanuele Trucco

Computing

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

7 Citations (Scopus)

Abstract

In this paper we propose a novel weakly-supervised feature learning approach, learning discriminative local features from image-level labelled data for image classification. Unlike existing feature learning approaches which assume that a set of additional data in the form of matching/non-matching pairs of local patches are given for learning the features, our approach only uses the image-level labels which are much easier to obtain. Experiments on a colonoscopy image dataset with 2100 images shows that the learned local features outperforms other hand-crafted features and gives a state-or-the-art classification accuracy of 93.5%.

Original language	English
Title of host publication	Proceedings - International Symposium on Biomedical Imaging
Publisher	IEEE Computer Society
Pages	420-423
Number of pages	4
Volume	2015-July
ISBN (Print)	9781479923748
DOIs	https://doi.org/10.1109/ISBI.2015.7163901
Publication status	Published - 2015
Event	12th IEEE International Symposium on Biomedical Imaging, ISBI 2015 - Brooklyn, United States Duration: 16 Apr 2015 → 19 Apr 2015

Conference

Conference	12th IEEE International Symposium on Biomedical Imaging, ISBI 2015
Country/Territory	United States
City	Brooklyn
Period	16/04/15 → 19/04/15

Keywords

Colonoscopy image classification
Discriminative feature learning
Local Binary Patterns

ASJC Scopus subject areas

Biomedical Engineering
Radiology Nuclear Medicine and imaging

Access to Document

10.1109/ISBI.2015.7163901

Cite this

@inproceedings{e9da44c94a744b7c9795e706b35bf37c,

title = "Learning discriminative local features from image-level labelled data for colonoscopy image classification",

abstract = "In this paper we propose a novel weakly-supervised feature learning approach, learning discriminative local features from image-level labelled data for image classification. Unlike existing feature learning approaches which assume that a set of additional data in the form of matching/non-matching pairs of local patches are given for learning the features, our approach only uses the image-level labels which are much easier to obtain. Experiments on a colonoscopy image dataset with 2100 images shows that the learned local features outperforms other hand-crafted features and gives a state-or-the-art classification accuracy of 93.5%.",

keywords = "Colonoscopy image classification, Discriminative feature learning, Local Binary Patterns",

author = "Siyamalan Manivannan and Emanuele Trucco",

year = "2015",

doi = "10.1109/ISBI.2015.7163901",

language = "English",

isbn = "9781479923748",

volume = "2015-July",

pages = "420--423",

booktitle = "Proceedings - International Symposium on Biomedical Imaging",

publisher = "IEEE Computer Society",

note = "12th IEEE International Symposium on Biomedical Imaging, ISBI 2015 ; Conference date: 16-04-2015 Through 19-04-2015",

}

Manivannan, S & Trucco, E 2015, Learning discriminative local features from image-level labelled data for colonoscopy image classification. in Proceedings - International Symposium on Biomedical Imaging. vol. 2015-July, 7163901, IEEE Computer Society, pp. 420-423, 12th IEEE International Symposium on Biomedical Imaging, ISBI 2015, Brooklyn, United States, 16/04/15. https://doi.org/10.1109/ISBI.2015.7163901

TY - GEN

T1 - Learning discriminative local features from image-level labelled data for colonoscopy image classification

AU - Manivannan, Siyamalan

AU - Trucco, Emanuele

PY - 2015

Y1 - 2015

N2 - In this paper we propose a novel weakly-supervised feature learning approach, learning discriminative local features from image-level labelled data for image classification. Unlike existing feature learning approaches which assume that a set of additional data in the form of matching/non-matching pairs of local patches are given for learning the features, our approach only uses the image-level labels which are much easier to obtain. Experiments on a colonoscopy image dataset with 2100 images shows that the learned local features outperforms other hand-crafted features and gives a state-or-the-art classification accuracy of 93.5%.

AB - In this paper we propose a novel weakly-supervised feature learning approach, learning discriminative local features from image-level labelled data for image classification. Unlike existing feature learning approaches which assume that a set of additional data in the form of matching/non-matching pairs of local patches are given for learning the features, our approach only uses the image-level labels which are much easier to obtain. Experiments on a colonoscopy image dataset with 2100 images shows that the learned local features outperforms other hand-crafted features and gives a state-or-the-art classification accuracy of 93.5%.

KW - Colonoscopy image classification

KW - Discriminative feature learning

KW - Local Binary Patterns

UR - http://www.scopus.com/inward/record.url?scp=84944324230&partnerID=8YFLogxK

U2 - 10.1109/ISBI.2015.7163901

DO - 10.1109/ISBI.2015.7163901

M3 - Conference contribution

AN - SCOPUS:84944324230

SN - 9781479923748

VL - 2015-July

SP - 420

EP - 423

BT - Proceedings - International Symposium on Biomedical Imaging

PB - IEEE Computer Society

T2 - 12th IEEE International Symposium on Biomedical Imaging, ISBI 2015

Y2 - 16 April 2015 through 19 April 2015

ER -

Learning discriminative local features from image-level labelled data for colonoscopy image classification

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Multi-modal Retinal Biomarkers for Vascular Dementia; Developing and Enabling Image Analysis Tools (Joint with University of Edinburgh)

Cite this

Learning discriminative local features from image-level labelled data for colonoscopy image classification

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Projects

Multi-modal Retinal Biomarkers for Vascular Dementia; Developing and Enabling Image Analysis Tools (Joint with University of Edinburgh)

Cite this