GA-based filter selection for representation in convolutional neural networks

Junbong Kim, Minki Lee, Jongeun Choi, Kisung Seo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

One of the deep learning models, a convolutional neural network (CNN) has been very successful in a variety of computer vision tasks. Features of a CNN are automatically generated, however, they can be further optimized since they often require large scale parallel operations and there exist the possibility of overlapping redundant features. The aim of this paper is to use feature selection via evolutionary algorithms to remove the irrelevant deep features. This will minimize the computational complexity and the amount of overfitting while maintaining a good quality of representation. We demonstrate the improvement of the filter representation by performing experiments on three data sets of CIFAR10, metal surface defects, and variation of MNIST and by analyzing the classification performance as well as the variance of the filter.

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2018 Workshops, Proceedings
EditorsLaura Leal-Taixé, Stefan Roth
PublisherSpringer Verlag
Pages609-618
Number of pages10
ISBN (Print)9783030110178
DOIs
Publication statusPublished - 2019 Jan 1
Event15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany
Duration: 2018 Sep 82018 Sep 14

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11132 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other15th European Conference on Computer Vision, ECCV 2018
CountryGermany
CityMunich
Period18/9/818/9/14

Fingerprint

Neural Networks
Filter
Neural networks
Surface Defects
Overfitting
Surface defects
Evolutionary algorithms
Computer Vision
Feature Selection
Computer vision
Overlapping
Evolutionary Algorithms
Feature extraction
Computational complexity
Computational Complexity
Metals
Minimise
Demonstrate
Experiment
Experiments

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Kim, J., Lee, M., Choi, J., & Seo, K. (2019). GA-based filter selection for representation in convolutional neural networks. In L. Leal-Taixé, & S. Roth (Eds.), Computer Vision – ECCV 2018 Workshops, Proceedings (pp. 609-618). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11132 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-11018-5_48
Kim, Junbong ; Lee, Minki ; Choi, Jongeun ; Seo, Kisung. / GA-based filter selection for representation in convolutional neural networks. Computer Vision – ECCV 2018 Workshops, Proceedings. editor / Laura Leal-Taixé ; Stefan Roth. Springer Verlag, 2019. pp. 609-618 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{bfa7c10dcaaf4684b2e8ca161758e52d,
title = "GA-based filter selection for representation in convolutional neural networks",
abstract = "One of the deep learning models, a convolutional neural network (CNN) has been very successful in a variety of computer vision tasks. Features of a CNN are automatically generated, however, they can be further optimized since they often require large scale parallel operations and there exist the possibility of overlapping redundant features. The aim of this paper is to use feature selection via evolutionary algorithms to remove the irrelevant deep features. This will minimize the computational complexity and the amount of overfitting while maintaining a good quality of representation. We demonstrate the improvement of the filter representation by performing experiments on three data sets of CIFAR10, metal surface defects, and variation of MNIST and by analyzing the classification performance as well as the variance of the filter.",
author = "Junbong Kim and Minki Lee and Jongeun Choi and Kisung Seo",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-11018-5_48",
language = "English",
isbn = "9783030110178",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "609--618",
editor = "Laura Leal-Taix{\'e} and Stefan Roth",
booktitle = "Computer Vision – ECCV 2018 Workshops, Proceedings",
address = "Germany",

}

Kim, J, Lee, M, Choi, J & Seo, K 2019, GA-based filter selection for representation in convolutional neural networks. in L Leal-Taixé & S Roth (eds), Computer Vision – ECCV 2018 Workshops, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11132 LNCS, Springer Verlag, pp. 609-618, 15th European Conference on Computer Vision, ECCV 2018, Munich, Germany, 18/9/8. https://doi.org/10.1007/978-3-030-11018-5_48

GA-based filter selection for representation in convolutional neural networks. / Kim, Junbong; Lee, Minki; Choi, Jongeun; Seo, Kisung.

Computer Vision – ECCV 2018 Workshops, Proceedings. ed. / Laura Leal-Taixé; Stefan Roth. Springer Verlag, 2019. p. 609-618 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11132 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - GA-based filter selection for representation in convolutional neural networks

AU - Kim, Junbong

AU - Lee, Minki

AU - Choi, Jongeun

AU - Seo, Kisung

PY - 2019/1/1

Y1 - 2019/1/1

N2 - One of the deep learning models, a convolutional neural network (CNN) has been very successful in a variety of computer vision tasks. Features of a CNN are automatically generated, however, they can be further optimized since they often require large scale parallel operations and there exist the possibility of overlapping redundant features. The aim of this paper is to use feature selection via evolutionary algorithms to remove the irrelevant deep features. This will minimize the computational complexity and the amount of overfitting while maintaining a good quality of representation. We demonstrate the improvement of the filter representation by performing experiments on three data sets of CIFAR10, metal surface defects, and variation of MNIST and by analyzing the classification performance as well as the variance of the filter.

AB - One of the deep learning models, a convolutional neural network (CNN) has been very successful in a variety of computer vision tasks. Features of a CNN are automatically generated, however, they can be further optimized since they often require large scale parallel operations and there exist the possibility of overlapping redundant features. The aim of this paper is to use feature selection via evolutionary algorithms to remove the irrelevant deep features. This will minimize the computational complexity and the amount of overfitting while maintaining a good quality of representation. We demonstrate the improvement of the filter representation by performing experiments on three data sets of CIFAR10, metal surface defects, and variation of MNIST and by analyzing the classification performance as well as the variance of the filter.

UR - http://www.scopus.com/inward/record.url?scp=85061737419&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85061737419&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-11018-5_48

DO - 10.1007/978-3-030-11018-5_48

M3 - Conference contribution

AN - SCOPUS:85061737419

SN - 9783030110178

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 609

EP - 618

BT - Computer Vision – ECCV 2018 Workshops, Proceedings

A2 - Leal-Taixé, Laura

A2 - Roth, Stefan

PB - Springer Verlag

ER -

Kim J, Lee M, Choi J, Seo K. GA-based filter selection for representation in convolutional neural networks. In Leal-Taixé L, Roth S, editors, Computer Vision – ECCV 2018 Workshops, Proceedings. Springer Verlag. 2019. p. 609-618. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-11018-5_48