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

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

    研究成果: Conference contribution

    5 被引用数 (Scopus)

    抄録

    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.

    本文言語English
    ホスト出版物のタイトルProceedings - 2018 1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018
    出版社Institute of Electrical and Electronics Engineers Inc.
    ページ27-30
    ページ数4
    ISBN(電子版)9781538692097
    DOI
    出版ステータスPublished - 2019 3 11
    イベント1st IEEE International Conference on Artificial Intelligence for Industries, AI4I 2018 - Laguna Hills, United States
    継続期間: 2018 9 262018 9 28

    出版物シリーズ

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

    Conference

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

    ASJC Scopus subject areas

    • 人工知能
    • 産業および生産工学

    フィンガープリント

    「Balanced mini-batch training for imbalanced image data classification with neural network」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

    引用スタイル