Regularized Pooling

Takato Otsuzuki, Hideaki Hayashi, Yuchen Zheng, Seiichi Uchida

研究成果: Chapter in Book/Report/Conference proceedingConference contribution

抄録

In convolutional neural networks (CNNs), pooling operations play important roles such as dimensionality reduction and deformation compensation. In general, max pooling, which is the most widely used operation for local pooling, is performed independently for each kernel. However, the deformation may be spatially smooth over the neighboring kernels. This means that max pooling is too flexible to compensate for actual deformations. In other words, its excessive flexibility risks canceling the essential spatial differences between classes. In this paper, we propose regularized pooling, which enables the value selection direction in the pooling operation to be spatially smooth across adjacent kernels so as to compensate only for actual deformations. The results of experiments on handwritten character images and texture images showed that regularized pooling not only improves recognition accuracy but also accelerates the convergence of learning compared with conventional pooling operations.

本文言語英語
ホスト出版物のタイトルArtificial Neural Networks and Machine Learning – ICANN 2020 - 29th International Conference on Artificial Neural Networks, Proceedings
編集者Igor Farkaš, Paolo Masulli, Stefan Wermter
出版社Springer Science and Business Media Deutschland GmbH
ページ241-254
ページ数14
ISBN(印刷版)9783030616151
DOI
出版ステータス出版済み - 2020
イベント29th International Conference on Artificial Neural Networks, ICANN 2020 - Bratislava, スロバキア
継続期間: 9 15 20209 18 2020

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
12397 LNCS
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

会議

会議29th International Conference on Artificial Neural Networks, ICANN 2020
Countryスロバキア
CityBratislava
Period9/15/209/18/20

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

フィンガープリント 「Regularized Pooling」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