Balanced mini-batch training for imbalanced image data classification with neural network

Ryota Shimizu, Kosuke Asako, Hiroki Ojima, Shohei Morinaga, Mototsugu Hamada, Tadahiro Kuroda

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

Abstract

We propose a novel method of training neural networks for industrial image classification that can reduce the effect of imbalanced data in supervised training. We considered visual quality inspection of industrial products as an image-classification task and attempted to solve this with a convolutional neural network; however, a problem of imbalanced data emerged in supervised training in which the neural network cannot optimize parameters. Since most industrial products are not defective, samples of defective products were fewer than those of the non-defective products; this difference in the number of samples causes an imbalance in training data. A neural network trained with imbalanced data often has varied levels of precision in determining each class depending on the difference in the number of class samples in the training data, which is a significant problem in industrial quality inspection. As a solution to this problem, we propose a balanced mini-batch training method that can virtually balance the class ratio of training samples. In an experiment, the neural network trained with the proposed method achieved higher classification ability than that trained with over-sampled or undersampled data for two types of imbalanced image datasets.

Original languageEnglish
Title of host publicationProceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages27-30
Number of pages4
ISBN (Electronic)9781538692097
DOIs
Publication statusPublished - 2019 Mar 11
Event1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018 - Laguna Hills, United States
Duration: 2018 Sep 262018 Sep 28

Publication series

NameProceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018

Conference

Conference1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018
CountryUnited States
CityLaguna Hills
Period18/9/2618/9/28

Fingerprint

Neural networks
Image classification
Inspection
Experiments

Keywords

  • convolutional neural network
  • image classification
  • imbalanced data
  • mini-batch training
  • visual quality inspection

ASJC Scopus subject areas

  • Artificial Intelligence
  • Industrial and Manufacturing Engineering

Cite this

Shimizu, R., Asako, K., Ojima, H., Morinaga, S., Hamada, M., & Kuroda, T. (2019). Balanced mini-batch training for imbalanced image data classification with neural network. In Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018 (pp. 27-30). [8665709] (Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/AI4I.2018.8665709

Balanced mini-batch training for imbalanced image data classification with neural network. / Shimizu, Ryota; Asako, Kosuke; Ojima, Hiroki; Morinaga, Shohei; Hamada, Mototsugu; Kuroda, Tadahiro.

Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018. Institute of Electrical and Electronics Engineers Inc., 2019. p. 27-30 8665709 (Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018).

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

Shimizu, R, Asako, K, Ojima, H, Morinaga, S, Hamada, M & Kuroda, T 2019, Balanced mini-batch training for imbalanced image data classification with neural network. in Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018., 8665709, Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018, Institute of Electrical and Electronics Engineers Inc., pp. 27-30, 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018, Laguna Hills, United States, 18/9/26. https://doi.org/10.1109/AI4I.2018.8665709
Shimizu R, Asako K, Ojima H, Morinaga S, Hamada M, Kuroda T. Balanced mini-batch training for imbalanced image data classification with neural network. In Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018. Institute of Electrical and Electronics Engineers Inc. 2019. p. 27-30. 8665709. (Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018). https://doi.org/10.1109/AI4I.2018.8665709
Shimizu, Ryota ; Asako, Kosuke ; Ojima, Hiroki ; Morinaga, Shohei ; Hamada, Mototsugu ; Kuroda, Tadahiro. / Balanced mini-batch training for imbalanced image data classification with neural network. Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 27-30 (Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018).
@inproceedings{59e392b55e5b4418a1983fa5094df2b8,
title = "Balanced mini-batch training for imbalanced image data classification with neural network",
abstract = "We propose a novel method of training neural networks for industrial image classification that can reduce the effect of imbalanced data in supervised training. We considered visual quality inspection of industrial products as an image-classification task and attempted to solve this with a convolutional neural network; however, a problem of imbalanced data emerged in supervised training in which the neural network cannot optimize parameters. Since most industrial products are not defective, samples of defective products were fewer than those of the non-defective products; this difference in the number of samples causes an imbalance in training data. A neural network trained with imbalanced data often has varied levels of precision in determining each class depending on the difference in the number of class samples in the training data, which is a significant problem in industrial quality inspection. As a solution to this problem, we propose a balanced mini-batch training method that can virtually balance the class ratio of training samples. In an experiment, the neural network trained with the proposed method achieved higher classification ability than that trained with over-sampled or undersampled data for two types of imbalanced image datasets.",
keywords = "convolutional neural network, image classification, imbalanced data, mini-batch training, visual quality inspection",
author = "Ryota Shimizu and Kosuke Asako and Hiroki Ojima and Shohei Morinaga and Mototsugu Hamada and Tadahiro Kuroda",
year = "2019",
month = "3",
day = "11",
doi = "10.1109/AI4I.2018.8665709",
language = "English",
series = "Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "27--30",
booktitle = "Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018",

}

TY - GEN

T1 - Balanced mini-batch training for imbalanced image data classification with neural network

AU - Shimizu, Ryota

AU - Asako, Kosuke

AU - Ojima, Hiroki

AU - Morinaga, Shohei

AU - Hamada, Mototsugu

AU - Kuroda, Tadahiro

PY - 2019/3/11

Y1 - 2019/3/11

N2 - We propose a novel method of training neural networks for industrial image classification that can reduce the effect of imbalanced data in supervised training. We considered visual quality inspection of industrial products as an image-classification task and attempted to solve this with a convolutional neural network; however, a problem of imbalanced data emerged in supervised training in which the neural network cannot optimize parameters. Since most industrial products are not defective, samples of defective products were fewer than those of the non-defective products; this difference in the number of samples causes an imbalance in training data. A neural network trained with imbalanced data often has varied levels of precision in determining each class depending on the difference in the number of class samples in the training data, which is a significant problem in industrial quality inspection. As a solution to this problem, we propose a balanced mini-batch training method that can virtually balance the class ratio of training samples. In an experiment, the neural network trained with the proposed method achieved higher classification ability than that trained with over-sampled or undersampled data for two types of imbalanced image datasets.

AB - We propose a novel method of training neural networks for industrial image classification that can reduce the effect of imbalanced data in supervised training. We considered visual quality inspection of industrial products as an image-classification task and attempted to solve this with a convolutional neural network; however, a problem of imbalanced data emerged in supervised training in which the neural network cannot optimize parameters. Since most industrial products are not defective, samples of defective products were fewer than those of the non-defective products; this difference in the number of samples causes an imbalance in training data. A neural network trained with imbalanced data often has varied levels of precision in determining each class depending on the difference in the number of class samples in the training data, which is a significant problem in industrial quality inspection. As a solution to this problem, we propose a balanced mini-batch training method that can virtually balance the class ratio of training samples. In an experiment, the neural network trained with the proposed method achieved higher classification ability than that trained with over-sampled or undersampled data for two types of imbalanced image datasets.

KW - convolutional neural network

KW - image classification

KW - imbalanced data

KW - mini-batch training

KW - visual quality inspection

UR - http://www.scopus.com/inward/record.url?scp=85064228342&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064228342&partnerID=8YFLogxK

U2 - 10.1109/AI4I.2018.8665709

DO - 10.1109/AI4I.2018.8665709

M3 - Conference contribution

AN - SCOPUS:85064228342

T3 - Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018

SP - 27

EP - 30

BT - Proceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -