https://github.com/97k/styler

Implementation of Image Style Transfer Using Convolutional Neural Networks, Paper by Gatys, Ecker, Bethge
https://github.com/97k/styler

deep-learning pytorch-implmention styletransfer vgg19

Last synced: 3 days ago
JSON representation

Implementation of Image Style Transfer Using Convolutional Neural Networks, Paper by Gatys, Ecker, Bethge

• Host: GitHub
• URL: https://github.com/97k/styler
• Owner: 97k
• Created: 2019-01-23T17:57:03.000Z (almost 6 years ago)
• Default Branch: master
• Last Pushed: 2019-01-25T06:18:05.000Z (almost 6 years ago)
• Last Synced: 2024-04-20T12:43:57.172Z (7 months ago)
• Topics: deep-learning, pytorch-implmention, styletransfer, vgg19
• Language: Jupyter Notebook
• Homepage: https://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Gatys_Image_Style_Transfer_CVPR_2016_paper.pdf
• Size: 2.59 MB
• Stars: 2
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Styler

Make Neural Net, paint for you!

Styler is a deep learning tool, in which style of one image is applied on style of another image!

#### It is based on transfer learning, and it is implementation of the research paper ["Image Style Transfer by CNN"](https://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Gatys_Image_Style_Transfer_CVPR_2016_paper.pdf)

I experimented it on Taj Mahal, and I'm attacing the styled version of it, that is output from the pretrained vgg19 after moulding it according to the paper.

## Result!

### Final Image, One that vgg19 painted!

![Taj Style](https://raw.githubusercontent.com/aditya98ak/Styler/master/assets/taj_style.png)

> This styled image is generated with help of a content and a style image.

### Content Image

![Content Image](https://raw.githubusercontent.com/aditya98ak/Styler/master/assets/taj_mahal.jpg)

### Style Image

![Style Image](https://raw.githubusercontent.com/aditya98ak/Styler/master/assets/wp3.jpg)

### Want to experiment?

### Clone

**1.)** Clone the repository by using this link :

```

$ git clone https://github.com/aditya98ak/Styler.git

```

### Experiment

**2.)** 

For better peformance upload the noeboook on [googlecolab](https://colab.research.google.com) and start experimenting, for example Change the style image, change the weights etc.

It's fun!

---

 Built with ♥ by Aditya Kaushik([@aditya98ak](https://github.com/aditya98ak))

 Thanks to [udacity](udacity.com)! For introducing me to Style Transfer.