deepdream.py

print('let us deep ream!')

from hack import hack
hack()

# aux functions

import scipy
from keras.preprocessing import image


def resize_img(img, size):
    img = np.copy(img)
    factors = (1,
               float(size[0]) / img.shape[1],
               float(size[1]) / img.shape[2],
               1)
    return scipy.ndimage.zoom(img, factors, order=1)


def save_img(img, fname):
    pil_img = deprocess_image(np.copy(img))
    scipy.misc.imsave(fname, pil_img)


def preprocess_image(image_path):
    img = image.load_img(image_path)
    img = image.img_to_array(img)
    img = np.expand_dims(img, axis=0)
    img = inception_v3.preprocess_input(img)
    return img


def deprocess_image(x):
    if K.image_data_format() == 'channels_first':
        x = x.reshape((3, x.shape[2], x.shape[3]))
        x = x.transpose((1, 2, 0))
    else:
        x = x.reshape((x.shape[1], x.shape[2], 3))
    x /= 2.
    x += 0.5
    x *= 255.
    x = np.clip(x, 0, 255).astype('uint8')
    return x

# the main code

from keras.applications import inception_v3
from keras import backend as K

K.set_learning_phase(0)

model = inception_v3.InceptionV3(weights='imagenet', include_top=False)

# model.summary()

layer_contributions = {
  'mixed2': 0.2, 
  'mixed3': 3.,
  'mixed4': 2.,
  'mixed5': 1.5,
}

layer_dict = dict([(layer.name, layer) for layer in model.layers])

loss = K.variable(0.)

for layer_name in layer_contributions:
    coeff = layer_contributions[layer_name]
    activation = layer_dict[layer_name].output

    scaling = K.prod(K.cast(K.shape(activation), 'float32'))
    loss = loss + coeff * K.sum(K.square(activation[:, 2: -2, 2: -2, :])) / scaling

dream = model.input

grads = K.gradients(loss, dream)[0]

grads /= K.maximum(K.mean(K.abs(grads)), 1e-7)

outputs = [loss, grads]

fetch_loss_and_grads = K.function([dream], outputs)

def eval_loss_and_grads(x): 
    outs = fetch_loss_and_grads([x])
    loss_value = outs[0]
    grad_values = outs[1]
    return loss_value, grad_values

def gradient_ascent(x, iterations, step, max_loss=None):
    for i in range(iterations):
        loss_value, grad_values = eval_loss_and_grads(x)
        if max_loss is not None and loss_value > max_loss:
            break
        print('... loss value at ', i, ':', loss_value)
        x += step * grad_values
    return x

import numpy as np 

step = 0.01
num_octave = 3
octave_scale = 1.4
iterations = 20

max_loss = 10.

base_img_path = '../bunny.jpg'

img = preprocess_image(base_img_path)

original_shape = img.shape[1:3]

successive_shapes = [original_shape]

for i in range(1, num_octave):
    shape = tuple([int(dim / (octave_scale ** i)) for dim in original_shape])
    successive_shapes.append(shape)

successive_shapes = successive_shapes[::-1]

original_image = np.copy(img)

shrunk_original_image = resize_img(img, successive_shapes[0])

for shape in successive_shapes:
    print('Processing shape: ', shape)
    img = resize_img(img, shape)
    img = gradient_ascent(img, iterations=iterations, step=step, max_loss=max_loss)

    upscaled_shrunk_original_img = resize_img(shrunk_original_image, shape)

    same_original_size = resize_img(shrunk_original_image, shape)

    lost_detail = same_original_size - upscaled_shrunk_original_img

    img += lost_detail

    shrunk_original_image = resize_img(original_image, shape)

    save_img(img, fname='../dream_at_scale_' + str(shape) + '.png')


save_img(img, fname='../final_dream.png')