AI-Art-Generator

Overview

Simple Art style transferer

You give a style image and the model learns the features and transfers it to the content image Best recommended to run with a GPU for fastest result

How it works?

Neural style transfer is an optimization technique used to take two images, style reference image (such as an artwork by a famous painter), and the input image you want to style

The model then blends the images to give your input image a artistic style

Further reading

Gatys’ paper
Gradient descent

Credits

Tensorflow article (It was for version Tensorflow V1)

Some examples of the project

Content and style images

Output

How to run locally?

Clone the repo.

git clone https://github.com/Rohith04MVK/AI-Art-Generator

Setup conda.

conda create -n AI-Art-Generator python=3.8.5

conda activate AI-Art-Generator

Install dependancies.

pip install -r requirement.txt

OR

conda install --file requirement.txt

Replace the pictures.

Replace line 10 content_path with the image you want to transform Replace line 11 style_path with the style picture you want to transfer

Run the file!

python aiart.py