Chinese Characters Style Analysis Using Generative Adversarial Network

DOI

10.2991/ncce-18.2018.153How to use a DOI?

Keywords

Chinese font; Character font;Convolutional neural network CNN; GAN.

Abstract

Various methods have been proposed in previous works to achieve effective printed Chinese character recognition. Feature extraction and production of large scale multi-font Chinese character remains a major challenge owing to the wide variety in the shape,layout,and grey-level distribution of single Chinese characters across different font styles.Convolutional neural networks (CNNs) have shown outstanding performances in many fields.Convolutional layer is the dominant algorithm used in training neural networks.In this paper,we propose a Generative Adversarial Network(GAN) to analyze chinese character fonts,extract font features through cnn,and we can output the type of fonts from the learning characters.There are many competitive CNN applications,aiming to achieve chinese font performance.In order to capture rich and discriminative information of fonts,we combine GAN with CNN to learn good features for the fonts and put out the desired fonts as required.The approach can generate more obvious font features and better display than original structure of GAN

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Haiyan Deng
AU  - Yijun Liu
PY  - 2018/05
DA  - 2018/05
TI  - Chinese Characters Style Analysis Using Generative Adversarial Network
BT  - Proceedings of the 2018 International Conference on Network, Communication, Computer Engineering (NCCE 2018)
PB  - Atlantis Press
SP  - 912
EP  - 916
SN  - 1951-6851
UR  - https://doi.org/10.2991/ncce-18.2018.153
DO  - 10.2991/ncce-18.2018.153
ID  - Deng2018/05
ER  -