Training convolutional autoencoders with metric learning

Yosuke Onitsuka, Wataru Ohyama, Seiichi Uchida

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

抄録

We propose a new Training method that enables an autoencoder to extract more useful features for retrieval or classification tasks with limited-size datasets. Some targets in document analysis and recognition (DAR) including signature verification, historical document analysis, and scene text recognition, involve a common problem in which the size of the dataset available for training is small against the intra-class variety of the target appearance. Recently, several approaches, such as variational autoencoders and deep metric learning, have been proposed to obtain a feature representation that is suitable for the tasks. However, these methods sometimes cause an overfitting problem in which the accuracy of the test data is relatively low, while the performance for the training dataset is quite high. Our proposed method obtains feature representations for such tasks in DAR using convolutional autoencoders with metric learning. The accuracy is evaluated on an image-based retrieval of ancient Japanese signatures.

本文言語英語
ホスト出版物のタイトルProceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019
出版社IEEE Computer Society
ページ86-91
ページ数6
ISBN(電子版)9781728128610
DOI
出版ステータス出版済み - 9 2019
イベント15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 - Sydney, オーストラリア
継続期間: 9 20 20199 25 2019

出版物シリーズ

名前Proceedings of the International Conference on Document Analysis and Recognition, ICDAR
ISSN(印刷版)1520-5363

会議

会議15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019
Countryオーストラリア
CitySydney
Period9/20/199/25/19

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

引用スタイル