Transferring styles to images

In the last couple of years, styles from one image to another has had an enormous boost thanks to deep learning. Many people have experimented with a certain style, often from a well-known painter, to a photo. The resulting images are often interesting to see because they show a mix between the painter's style and the original image. In the following recipe, we will show you how to use pretrained weights from VGG16 to transfer the style of one image to another. 

How to do it...

1. We start importing all the libraries as follows:

from keras.preprocessing.image import load_img, img_to_array
from scipy.misc import imsave
import numpy as np
from scipy.optimize import fmin_l_bfgs_b
import time
import argparse

from keras.applications import vgg16
from keras import backend as K

2. Next, we load the two images that we will use for style transfer and plot them:

base_image_path = 'Data/golden_gate.jpg'
style_reference_image_path = 'Data/starry_night.jpg'
result_prefix = 'result_...