-
Notifications
You must be signed in to change notification settings - Fork 0
/
styleTransfer.py
184 lines (160 loc) · 6.52 KB
/
styleTransfer.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
# Code from https://github.com/walid0925/AI_Artistry/blob/master/main.py
import numpy as np
import pandas as pd
import tensorflow as tf
from PIL import Image
from keras import backend as K
from keras.preprocessing.image import load_img, img_to_array
from keras.applications import VGG16
from keras.applications.vgg16 import preprocess_input
from keras.layers import Input
from scipy.optimize import fmin_l_bfgs_b
import time
import os
# Select GPU and limit memory usage
os.environ["CUDA_VISIBLE_DEVICES"]="0"
gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=.33)
sess = tf.Session(config=tf.ConfigProto(gpu_options=gpu_options))
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~##
## Specify paths for 1) content image 2) style image
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~##
cImPath = 'dataset/training_set/dogs/dog.175.jpg'
sImPath = []
sImPath.append('Styles/tsunami.jpg')
sImPath.append('Styles/the_scream.jpg')
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~##
## Image processing
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~##
targetHeight = 512
targetWidth = 512
targetSize = (targetHeight, targetWidth)
cImageOrig = Image.open(cImPath)
cImageSizeOrig = cImageOrig.size
cImage = load_img(path=cImPath, target_size=targetSize)
cImArr = img_to_array(cImage)
cImArr = K.variable(preprocess_input(np.expand_dims(cImArr, axis=0)), dtype='float32')
sImage = []
sImArr = []
for i in range(len(sImPath)):
sImage.append(load_img(path=sImPath[i], target_size=targetSize))
sImArr.append(img_to_array(sImage[i]))
sImArr[i] = K.variable(preprocess_input(np.expand_dims(sImArr[i], axis=0)), dtype='float32')
gIm0 = np.random.randint(256, size=(targetWidth, targetHeight, 3)).astype('float64')
gIm0 = preprocess_input(np.expand_dims(gIm0, axis=0))
gImPlaceholder = K.placeholder(dtype="float",shape=(1, targetWidth, targetHeight, 3))
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~##
## Define loss and helper functions
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~##
def get_feature_reps(x, layer_names, model):
featMatrices = []
for ln in layer_names:
selectedLayer = model.get_layer(ln)
featRaw = selectedLayer.output
featRawShape = K.shape(featRaw).eval(session=tf_session)
N_l = featRawShape[-1]
M_l = featRawShape[1]*featRawShape[2]
featMatrix = K.reshape(featRaw, (M_l, N_l))
featMatrix = K.transpose(featMatrix)
featMatrices.append(featMatrix)
return featMatrices
def get_content_loss(F, P):
cLoss = 0.5*K.sum(K.square(F - P))
return cLoss
def get_Gram_matrix(F):
G = K.dot(F, K.transpose(F))
return G
def get_style_loss(ws, Gs, As):
sLoss = K.variable(0.)
for w, G, A in zip(ws, Gs, As):
M_l = K.int_shape(G)[1]
N_l = K.int_shape(G)[0]
G_gram = get_Gram_matrix(G)
A_gram = get_Gram_matrix(A)
sLoss = sLoss + w*0.25*K.sum(K.square(G_gram - A_gram))/ (N_l**2 * M_l**2)
return sLoss
def get_total_loss(gImPlaceholder, alpha=1.0, beta=10000.0):
F = get_feature_reps(gImPlaceholder, layer_names=[cLayerName], model=gModel)[0]
Gs = get_feature_reps(gImPlaceholder, layer_names=sLayerNames, model=gModel)
contentLoss = get_content_loss(F, P)
styleLossArr = []
styleLoss = 0
for i in range(len(sImPath)):
styleLossArr.append(get_style_loss(ws, Gs, As[i]))
styleLoss = styleLoss + styleLossArr[i]
totalLoss = alpha*contentLoss + beta*styleLoss
return totalLoss
def calculate_loss(gImArr):
"""
Calculate total loss using K.function
"""
if gImArr.shape != (1, targetWidth, targetWidth, 3):
gImArr = gImArr.reshape((1, targetWidth, targetHeight, 3))
loss_fcn = K.function([gModel.input], [get_total_loss(gModel.input)])
return loss_fcn([gImArr])[0].astype('float64')
def get_grad(gImArr):
"""
Calculate the gradient of the loss function with respect to the generated image
"""
if gImArr.shape != (1, targetWidth, targetHeight, 3):
gImArr = gImArr.reshape((1, targetWidth, targetHeight, 3))
grad_fcn = K.function([gModel.input], K.gradients(get_total_loss(gModel.input), [gModel.input]))
grad = grad_fcn([gImArr])[0].flatten().astype('float64')
return grad
def postprocess_array(x):
# Zero-center by mean pixel
if x.shape != (targetWidth, targetHeight, 3):
x = x.reshape((targetWidth, targetHeight, 3))
x[..., 0] = x[..., 0] + 103.939
x[..., 1] = x[..., 1] + 116.779
x[..., 2] = x[..., 2] + 123.68
# 'BGR'->'RGB'
x = x[..., ::-1]
x = np.clip(x, 0, 255)
x = x.astype('uint8')
return x
def reprocess_array(x):
x = np.expand_dims(x.astype('float64'), axis=0)
x = preprocess_input(x)
return x
def save_original_size(x, iter, target_size=cImageSizeOrig):
xIm = Image.fromarray(x)
xIm = xIm.resize(target_size)
xIm.save('Results/resultTsunamiScreamDog%d.jpg' % (iter))
return xIm
tf_session = K.get_session()
cModel = VGG16(include_top=False, weights='imagenet', input_tensor=cImArr)
sModel = []
for i in range(len(sImPath)):
sModel.append(VGG16(include_top=False, weights='imagenet', input_tensor=sImArr[i]))
gModel = VGG16(include_top=False, weights='imagenet', input_tensor=gImPlaceholder)
cLayerName = 'block4_conv2'
sLayerNames = [
'block1_conv1',
'block2_conv1',
'block3_conv1',
'block4_conv1',
#'block5_conv1'
]
P = get_feature_reps(x=cImArr, layer_names=[cLayerName], model=cModel)[0]
As = []
for i in range(len(sImPath)):
As.append(get_feature_reps(x=sImArr[i], layer_names=sLayerNames, model=sModel[i]))
ws = np.ones(len(sLayerNames))/float(len(sLayerNames))
# Allows saving images at increasing runtimes
iterations = 1
base = 200
x_val = gIm0.flatten()
for iteration in range(1, iterations+1):
start = time.time()
xopt, f_val, info= fmin_l_bfgs_b(calculate_loss, x_val, fprime=get_grad,
maxiter=iteration * base, disp=True)
xOut = postprocess_array(xopt)
xIm = save_original_size(xOut, iteration*base)
print('Image saved')
end = time.time()
print('Time taken: {}'.format(end-start))
# https://stackoverflow.com/questions/10640114/overlay-two-same-sized-images-in-python
background = Image.open('Results/resultTsunamiScreamDog%d.jpg' % (iterations * base))
foreground = Image.open('dogTransparent.png')
background.paste(foreground, (0,0), foreground)
background.save('ArtDog.png', 'PNG')