Japanese kanji font style transfer based on GAN with unpaired training

Hiroki Sakai, Daisuke Niino, Takashi Ijiri

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

To design a whole package of Japanese font is labor consuming, since it usually contains about 30k kanji characters. To support an efficient design process, this poster attempts to adopt a style transfer algorithm for font package completion. Given two font packages where one contains all characters and the other lacks a large part, we train CycleGAN to perform style transfer between the two packages and transfer the style from the former to the latter. To illustrate the feasibility of our technique, we performed style transfer experiments and achieved visually plausible results by using a relatively small training data set.

Original languageEnglish
Title of host publicationPacific Graphics 2018 - 26th Pacific Conference on Computer Graphics and Applications, Short Papers and Posters Proceedings
EditorsHongbo Fu, Abhijeet Ghosh, Johannes Kopf
PublisherIEEE Computer Society
Pages5-6
Number of pages2
ISBN (Electronic)9783038680734
DOIs
Publication statusPublished - 2018
Event26th Pacific Conference on Computer Graphics and Applications, Pacific Graphics 2018 - Hong Kong, Hong Kong
Duration: 2018 Oct 82018 Oct 11

Publication series

NameProceedings - Pacific Conference on Computer Graphics and Applications
Volume2018-October
ISSN (Print)1550-4085

Conference

Conference26th Pacific Conference on Computer Graphics and Applications, Pacific Graphics 2018
Country/TerritoryHong Kong
CityHong Kong
Period18/10/818/10/11

ASJC Scopus subject areas

  • Software
  • Computer Graphics and Computer-Aided Design
  • Modelling and Simulation

Fingerprint

Dive into the research topics of 'Japanese kanji font style transfer based on GAN with unpaired training'. Together they form a unique fingerprint.

Cite this