Recognition model based on residual networks for cursive Hanja recognition

Sangwon Lee, Gil Jin Jang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations

Abstract

With the development of algorithms and computer skills, deep learning using CNN (convolutional neural network) has been applied to various fields, especially in image processing field. In this paper, we designed an improved model based on ResNet with CNN structure, and learned the database. The Chaucer Database used in the experiment consisted of 824 Chinese characters among the Chinese characters registered in the Dictionary of Old Document Type Usage Dictionary of the Korean Studies Data Portal and total of 240,000 data. The experiment used 10-fold cross-validation. The ResNet-based percussive network used in the experiments showed an average of 94.7% top-1 accuracy in the post hoc classification test.

Original languageEnglish
Title of host publicationInternational Conference on Information and Communication Technology Convergence
Subtitle of host publicationICT Convergence Technologies Leading the Fourth Industrial Revolution, ICTC 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages579-583
Number of pages5
ISBN (Electronic)9781509040315
DOIs
StatePublished - 12 Dec 2017
Event8th International Conference on Information and Communication Technology Convergence, ICTC 2017 - Jeju Island, Korea, Republic of
Duration: 18 Oct 201720 Oct 2017

Publication series

NameInternational Conference on Information and Communication Technology Convergence: ICT Convergence Technologies Leading the Fourth Industrial Revolution, ICTC 2017
Volume2017-December

Conference

Conference8th International Conference on Information and Communication Technology Convergence, ICTC 2017
Country/TerritoryKorea, Republic of
CityJeju Island
Period18/10/1720/10/17

Keywords

  • Convolutional neural network
  • Hanja
  • OCR (optical character recognition)
  • residual network

Fingerprint

Dive into the research topics of 'Recognition model based on residual networks for cursive Hanja recognition'. Together they form a unique fingerprint.

Cite this