ChineseCalligraphyGenerator

KO | EN

gitlite — search

#react #golang #android #machine-learning #docker #macos #java #windows #awesome-list #vue #pytorch #security #mcp #openai

ChineseCalligraphyGenerator

(★ 19)

ChineseCalligraphyGenerator

ChineseCalligraphyGenerator 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# GoogleML-Calligraphy

* [Overview](#overview)
* [OurCalligraphy](#OurCalligraphy)
* [Result](#Result)
* [Reference](#Reference)

## overview
Picture Style-Transfer is a tough problem in Computer Vision, which attracts lots attention in the recent decade. Different with the previous Style-Transfer for scenery pictures, we focus more on the Style-Transfer for Chinese Calligraphy.

URL:http://holer.cc:50378/calligraphy/

![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/demo1.jpg)
## OurCalligraphy
To solve the problem probably, we design a GAN-based model, which consists of two parts, a U-Net based Generator and a CNN-based Discriminator. 
Moreover, to enable multi-style transfer, we concatenate the encoded feature with embedding of the target style, and feed the mixed feature into decoder.
To enhanced the capability of generator in the in-domain dataset (etc. Lanting Jixu), we first train our model by adversarial learning in the out-domain dataset, which consist of nearly 50k Chinese Calligraphy pictures with different styles and fine-tune in a in-domain dataset.

### Network Structure
![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/network.jpg)

To display the generated result visually, we build a Chinese Calligraphy Generating (CCG) System., which is mainly based on Django, which enable users to generate specific style of Chinese Calligraphy for a certain text.
To extend the application of our Chinese Calligraph Generating System, we even train a auto couplet generator based on Transformer, which could given a reasonable couplet, paired to the given one. 
CCG will transfer the generated couplets into a certain style of  Chinese Calligraph, just like how human do in real world.

## Result
The reconstruction result of Lanting Jixu generated by our model:

![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/result.jpg)

Couplets generated by our model:

![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/couplet2.png)
![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/couplet3.png)
![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/couplet4.png)

Here are some system screenshots

![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/Interface1.png)

![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/Interface2.png)

![alt network](https://raw.githubusercontent.com/iamsohard/ChineseCalligraphyGenerator/master/assets/Interface3.png)

## Reference

* [zi2zi](https://github.com/pcgreat/zi2zi) by [pcgreat](https://github.com/pcgreat)
* [AGIS-Net](https://github.com/hologerry/AGIS-Net) by [hologerry](https://github.com/hologerry)
* [HCCG-CycleGAN](https://github.com/changebo/HCCG-CycleGAN) by [changebo](https://github.com/changebo)

* [pix2pix-tensorflow](https://github.com/yenchenlin/pix2pix-tensorflow) by [yenchenlin](https://github.com/yenchenlin)
* [Domain Transfer Network](https://github.com/yunjey/domain-transfer-network) by [yunjey](https://github.com/yunjey)
* [ac-gan](https://github.com/buriburisuri/ac-gan) by [buriburisuri](https://github.com/buriburisuri)
* [dc-gan](https://github.com/carpedm20/DCGAN-tensorflow) by [carpedm20](https://github.com/carpedm20)
* [origianl pix2pix torch code](https://github.com/phillipi/pix2pix) by [phillipi](https://github.com/phillipi)

[1] Isola P, Zhu J, Zhou T, et al. Image-to-Image Translation with Conditional Adversarial Networks[C]. computer vision and pattern recognition, 2017: 5967-5976.

[2] Odena A, Olah C, Shlens J, et al. Conditional Image Synthesis With Auxiliary Classifier GANs[J]. arXiv: Machine Learning, 2016.

[3] Taigman Y, Polyak A, Wolf L, et al. Unsupervised Cross-Domain Image Generation[J]. arXiv: Computer Vision and Pattern Recognition, 2016.

iamsohard

ChineseCalligraphyGenerator

ChineseCalligraphyGenerator 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

ChineseCalligraphyGenerator

ChineseCalligraphyGenerator 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README