Skip to content

Latest commit

 

History

History
33 lines (26 loc) · 1.18 KB

README.md

File metadata and controls

33 lines (26 loc) · 1.18 KB

Neural Style Transfer

This repository uses VGG-16 or VGG-19 pretrained model to do Neural Style transfer. Download the checkpoint file for VGG-16 here and for VGG-19 here

Usage

train.py [-h] [--cnt cnt_img] [--stl stl_img] [--size size [size ...]]
                [--model model] [--niters n_iters]

optional arguments:
  -h, --help            show this help message and exit
  --cnt cnt_img         Content Image for style transfer
  --stl stl_img         Style Image for style transfer
  --size size [size ...]
                        Size of output image [height, width]
  --model model         Model type to use ('vgg16' or 'vgg19'). default -
                        vgg19
  --niters n_iters      Number of iterations to run for

Dependencies

  1. Numpy
  2. Scipy
  3. Tensorflow

Content Image

Style Image

Output Image