Skip to content

Neural Style Transfer For Chinese Characters, implemented in Tensorflow 2, with custom dataset.

Notifications You must be signed in to change notification settings

huangxinping/CalliGAN-tf-keras

Repository files navigation

CalliGAN - Style and structure-aware Chinese Calligraphy Generator

CalliGAN - Tensorflow 2 Implementation (AI for Content Creation Workshop CVPR 2020)

output

Intruduction

This is the implementation of paper "CalliGAN - Style and structure-aware Chinese Calligraphy Generator" accepted by AI for Content Creation Workshop CVPR 2020.
The purpose of this paper is to generate Chinese calligraphy characters.

Environments

Tensorflow: 2.9.1
Keras: 2.9.0
Datasets: Custom, please refer to the "datasets" folder or 書法教學資料庫.

References

Paper: CalliGAN: Style and Structure-aware Chinese Calligraphy Character Generator
Code: An official implementation
Rewrite: Rewrite-tf-keras - An unofficial implementation, which is also to generate Chinese calligraphy cahracters
XingZi: XingZi-tf-keras - An unofficial implementation, which is also to generate Chinese calligraphy cahracters
Typography Research Collection: https://hackmd.io/@ishengfang/B1BZCZpPX?type=view

Disclaimer

An unofficail implementation, please refer to the original paper.

About

Neural Style Transfer For Chinese Characters, implemented in Tensorflow 2, with custom dataset.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published