/NeuralStyleTransfer

Generating new image out of 2 separate images by taking content of one image and style of other, using Neural Networks

Primary LanguageJupyter Notebook

Art Generation using Neural Style Transfer

Inputs a content image and a style image. Takes content of content image and generates an image with that using Neural Networks, but with the style extracted from style image.

image

Requirements

  • Tensorflow < 1.12
  • Python 3.5+
  • MatplotLib
  • Imageio
  • Numpy

Downloading pre-trained Model:

Click Here to download pre-trained VGG model used in order to generate the images.

To Run:

  • Download the pretrained model and place it in a folder named "pretrained-model"
  • Run the Jupyter Notebook

Work in Progress:

  • Web implementation for easy access

References:

  • Neural Style Transfer by Andrew Ng