Comparative study of deep learning methods for one-shot image classification (abstract)

J. van den Bogaert, H. Mohseni, Mahmoud Khodier, Yuliyan Stoyanov, D.C. Mocanu, V. Menkovski

Onderzoeksoutput: Bijdrage aan congresAbstract

196 Downloads (Pure)

Uittreksel

Training deep learning models for images classification requires large amount of labeled data to overcome the challenges of overfitting and underfitting. Usually, in many practical applications, these labeled data are not available. In an attempt to solve this problem, the one-shot learning paradigm tries to create machine learning models capable to learn well from one or (maximum) few labeled examples per class. To understand better the behavior of various deep learning models and approaches for one-shot learning, in this abstract, we perform a comparative study of the most used ones, on a challenging real-world dataset, i.e Fashion-MNIST.
Originele taal-2Engels
StatusGepubliceerd - 1 dec 2017
EvenementDutch-Belgian Database Day 2017 (DBDBD 2017) - Eindhoven, Nederland
Duur: 1 dec 20171 dec 2017

Workshop

WorkshopDutch-Belgian Database Day 2017 (DBDBD 2017)
LandNederland
StadEindhoven
Periode1/12/171/12/17

Vingerafdruk

Image classification
Learning systems
Deep learning

Citeer dit

van den Bogaert, J., Mohseni, H., Khodier, M., Stoyanov, Y., Mocanu, D. C., & Menkovski, V. (2017). Comparative study of deep learning methods for one-shot image classification (abstract). Abstract van Dutch-Belgian Database Day 2017 (DBDBD 2017), Eindhoven, Nederland.
van den Bogaert, J. ; Mohseni, H. ; Khodier, Mahmoud ; Stoyanov, Yuliyan ; Mocanu, D.C. ; Menkovski, V. / Comparative study of deep learning methods for one-shot image classification (abstract). Abstract van Dutch-Belgian Database Day 2017 (DBDBD 2017), Eindhoven, Nederland.
@conference{9125fe7822384bec929c7ab9fbad7ec1,
title = "Comparative study of deep learning methods for one-shot image classification (abstract)",
abstract = "Training deep learning models for images classification requires large amount of labeled data to overcome the challenges of overfitting and underfitting. Usually, in many practical applications, these labeled data are not available. In an attempt to solve this problem, the one-shot learning paradigm tries to create machine learning models capable to learn well from one or (maximum) few labeled examples per class. To understand better the behavior of various deep learning models and approaches for one-shot learning, in this abstract, we perform a comparative study of the most used ones, on a challenging real-world dataset, i.e Fashion-MNIST.",
author = "{van den Bogaert}, J. and H. Mohseni and Mahmoud Khodier and Yuliyan Stoyanov and D.C. Mocanu and V. Menkovski",
year = "2017",
month = "12",
day = "1",
language = "English",
note = "Dutch-Belgian Database Day 2017 (DBDBD 2017) ; Conference date: 01-12-2017 Through 01-12-2017",

}

van den Bogaert, J, Mohseni, H, Khodier, M, Stoyanov, Y, Mocanu, DC & Menkovski, V 2017, 'Comparative study of deep learning methods for one-shot image classification (abstract)', Dutch-Belgian Database Day 2017 (DBDBD 2017), Eindhoven, Nederland, 1/12/17 - 1/12/17.

Comparative study of deep learning methods for one-shot image classification (abstract). / van den Bogaert, J.; Mohseni, H.; Khodier, Mahmoud; Stoyanov, Yuliyan; Mocanu, D.C.; Menkovski, V.

2017. Abstract van Dutch-Belgian Database Day 2017 (DBDBD 2017), Eindhoven, Nederland.

Onderzoeksoutput: Bijdrage aan congresAbstract

TY - CONF

T1 - Comparative study of deep learning methods for one-shot image classification (abstract)

AU - van den Bogaert, J.

AU - Mohseni, H.

AU - Khodier, Mahmoud

AU - Stoyanov, Yuliyan

AU - Mocanu, D.C.

AU - Menkovski, V.

PY - 2017/12/1

Y1 - 2017/12/1

N2 - Training deep learning models for images classification requires large amount of labeled data to overcome the challenges of overfitting and underfitting. Usually, in many practical applications, these labeled data are not available. In an attempt to solve this problem, the one-shot learning paradigm tries to create machine learning models capable to learn well from one or (maximum) few labeled examples per class. To understand better the behavior of various deep learning models and approaches for one-shot learning, in this abstract, we perform a comparative study of the most used ones, on a challenging real-world dataset, i.e Fashion-MNIST.

AB - Training deep learning models for images classification requires large amount of labeled data to overcome the challenges of overfitting and underfitting. Usually, in many practical applications, these labeled data are not available. In an attempt to solve this problem, the one-shot learning paradigm tries to create machine learning models capable to learn well from one or (maximum) few labeled examples per class. To understand better the behavior of various deep learning models and approaches for one-shot learning, in this abstract, we perform a comparative study of the most used ones, on a challenging real-world dataset, i.e Fashion-MNIST.

M3 - Abstract

ER -

van den Bogaert J, Mohseni H, Khodier M, Stoyanov Y, Mocanu DC, Menkovski V. Comparative study of deep learning methods for one-shot image classification (abstract). 2017. Abstract van Dutch-Belgian Database Day 2017 (DBDBD 2017), Eindhoven, Nederland.