分布の形状を考慮した摂動に基づくデータ拡張 ― 敵対的生成ネットワークを用いた応用 ―

カテゴリ: 論文誌(論文単位)

グループ名: 【C】電子・情報・システム部門

発行日: 2019/12/01

タイトル(英語): Perturbation-induced Data Augmentation Considering Shape of Distribution ― An Application of Generative Adversarial Networks ―

著者名: 但馬　慶行（（株）日立製作所　研究開発グループ），河野　洋平（（株）日立製作所　研究開発グループ）

著者名(英語): Yoshiyuki Tajima (Hitachi Co., Ltd. Research and Development group), Yohei Kono (Hitachi Co., Ltd. Research and Development group)

キーワード: データ拡張，過正則化，敵対的生成ネットワーク，深層学習，摂動　　data augmentation，over regularization，generative adversarial networks，deep learning，perturbation

要約(英語): Machine Learning (ML) techniques need a tremendous volume of training data. However, in operation and maintenance of industrial facilities, it is difficult to get such a volume of data due to the lack of sensors or infrequency of target phenomena. A promising approach to solve this problem is so-called data augmentation, which generates training data by using a prior knowledge or adding noise (perturbation) to original data. For this, Gaussian noise is generally used because of its simplicity. However, when the distribution of original data is not isotropic, the Gaussian-based augmentation breaks its shape, which causes so-called over regularization. In this paper, we propose a novel perturbation-induced data augmentation method, which does not require any prior knowledge and makes it easy to control the magnitude of perturbation. The novelty of the proposed method is to keep certain characteristics of shape of original distribution. The perturbation for this is generated by combined use of generative adversarial networks and newly proposed objective functions. We experimentally show that the proposed method enables to keep the gap between peaks of a mixed normal distribution. The effectiveness of the proposed method is also demonstrated in the case of an image classification task.

本誌: 電気学会論文誌C（電子・情報・システム部門誌） Vol.139 No.12 （2019） 特集：電気・電子・情報関係学会東海支部連合大会

本誌掲載ページ: 1501-1508 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/139/12/139_1501/_article/-char/ja/