forked from Tobias-Fischer/rt_gene
-
Notifications
You must be signed in to change notification settings - Fork 0
/
GlassesCompletion.py
executable file
·202 lines (146 loc) · 7.63 KB
/
GlassesCompletion.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
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
from __future__ import print_function, division, absolute_import
from models import LSGAN_Model, Completion_Model
from utils import *
import os
from glob import glob
from tqdm import tqdm
from tensorflow.keras import backend as K
import external.poissonblending as blending
class GlassesCompletion(object):
def __init__(self, dataset_common_folder_path, dataset):
epoch_num_total = {
"s000": 100,
"s001": 100,
"s002": 86,
"s003": 100,
"s004": 76,
"s005": 77,
"s006": 100,
"s007": 81,
"s008": 73,
"s009": 58,
"s010": 70,
"s011": 53,
"s012": 99,
"s013": 29,
"s014": 74,
"s015": 98,
"s016": 95
}
epoch_num = epoch_num_total[dataset]
self.folder_path_images = dataset_common_folder_path + dataset + "_glasses"
self.path_images = self.folder_path_images + "/original/face_before_inpainting"
self.path_completion = self.folder_path_images + "/original/inpainting"
self.path_GAN_model = dataset_common_folder_path + dataset + "_noglasses/natural/GAN/model"
if not os.path.exists(self.path_completion):
os.mkdir(self.path_completion)
if not os.path.exists(self.path_completion+'/hats'):
os.mkdir(self.path_completion+'/hats')
if not os.path.exists(self.path_completion+'/blended'):
os.mkdir(self.path_completion+'/blended')
self.img_rows = 224
self.img_cols = 224
self.channel = 3
self.image_shape = [self.img_rows, self.img_cols, self.channel]
self.dataset = dataset
# For GAN
self.noise_dim = 100
self.generator = tf.keras.models.load_model(self.path_GAN_model+"/GAN_"+str(epoch_num)+"_"+self.dataset+"_forganECCV_generator_uniform.h5")
self.adversarial_cost = tf.keras.models.load_model(self.path_GAN_model+"/GAN_"+str(epoch_num)+"_"+self.dataset+"_forganECCV_adversarial_model_uniform.h5", custom_objects={'loss_LSGAN': loss_LSGAN})
print('Done Loading Pre-trained Network!')
def image_completion_random_search(self, nIter=1000, GPU_ID="0"):
filename_total_face = sorted(glob(os.path.join(self.path_images, 'face_*.png')))
num_total_data = len(filename_total_face)
print(num_total_data)
print('=======================================================')
GAN_4_loss = LSGAN_Model(self.img_rows, self.img_cols, self.channel, self.noise_dim, self.dataset)
dis_4_loss = GAN_4_loss.discriminator()
gen_4_loss = GAN_4_loss.generator()
GAN_Completion_model = Completion_Model(self.noise_dim)
complete_loss_model = GAN_Completion_model.cal_complete_loss(gen_4_loss, dis_4_loss)
mask_tensor = tf.compat.v1.placeholder(tf.float32, self.image_shape, name='mask')
images_tensor = tf.compat.v1.placeholder(tf.float32, self.image_shape, name='real_images')
G_images_tensor = tf.compat.v1.placeholder(tf.float32, self.image_shape, name='fake_images')
loss_contextual_temp = tf.abs(tf.multiply(mask_tensor, G_images_tensor) - tf.multiply(mask_tensor, images_tensor))
loss_contextual = tf.reduce_sum(input_tensor=tf.reshape(loss_contextual_temp, [tf.shape(input=loss_contextual_temp)[0], -1]), axis=1)
loss_perceptual = complete_loss_model.output[0]
loss = loss_contextual + 0.1*loss_perceptual
gradients = K.gradients(loss, complete_loss_model.input)
print('gradients: ', gradients)
print('=======================================================')
config = tf.compat.v1.ConfigProto()
config.gpu_options.visible_device_list = GPU_ID
config.gpu_options.allow_growth = True
sess = tf.compat.v1.Session(config=config)
sess.run(tf.compat.v1.global_variables_initializer())
print(self.path_completion)
for img_idx in tqdm(range(0, num_total_data)):
filename_face = filename_total_face[img_idx]
filename_index = filename_face[-14:-8]
filename_mask = self.folder_path_images + '/original/mask/mask_' + filename_index + '_overlay.png'
filename_out = self.path_completion+'/blended/' + filename_index + '.png'
if os.path.isfile(filename_out):
continue
data_face = imread_PRL(filename_face, is_grayscale=False)
image_face = np.array(data_face).astype(np.float32)
data_mask = imread_PRL(filename_mask, is_grayscale=True)
image_mask = np.array(data_mask).astype(np.float32)
# Sample index
sample_num = 1
# sample_noise_input = np.random.uniform(-1.0, 1.0, size=[sample_num, self.noise_dim])
# mask generation
mask = self.mask_PRL_Glasses(image_mask)
# masked_images = np.multiply(image_face, mask)
# y = np.ones([sample_num, 1])
zhats = np.random.uniform(-1.0, 1.0, size=[sample_num, self.noise_dim])
# loss_buf = 0
l_buf = 10000000
zhats_buf = zhats
# final_iter = 0
for j in range(nIter):
zhats_search = np.random.uniform(-1.0, 1.0, size=[sample_num, self.noise_dim])
G_imgs = self.generator.predict(zhats_search)
G_imgs = np.squeeze(G_imgs)
g, l, lc, lp = sess.run([gradients, loss, loss_contextual, loss_perceptual], feed_dict={complete_loss_model.input: zhats_search, mask_tensor: mask, images_tensor: image_face, G_images_tensor: G_imgs})
if np.sum(l) < l_buf:
l_buf = np.sum(l)
zhats_buf = zhats_search
# final_iter = j
zhats = zhats_buf
G_imgs = self.generator.predict(zhats)
G_imgs = np.squeeze(G_imgs)
# --------------------------------------------------------------
# Generate completed images
# inv_masked_hat_images = np.multiply(G_imgs, 1.0-mask)
# completed = masked_images + inv_masked_hat_images
filename = self.path_completion+'/hats/' + filename_index + '.png'
scipy.misc.imsave(filename, (G_imgs + 1) / 2)
# Poisson Blending
image_out = self.iminvtransform(G_imgs)
image_in = self.iminvtransform(image_face)
try:
image_out = self.poissonblending(image_in, image_out, mask)
filename = self.path_completion+'/blended/' + filename_index + '.png'
scipy.misc.imsave(filename, image_out)
except:
print("Error occurred while blending: " + str(filename_index))
pass
sess.close()
def mask_PRL_Glasses(self, mask_images):
mask = np.ones(self.image_shape)
for ir in range(self.img_rows):
for ic in range(self.img_cols):
# if mask_images[ir,ic] >= (127.5/127.5-1):
if mask_images[ir, ic] > (0/127.5-1):
mask[ir, ic, :] = 0
return mask
@staticmethod
def poissonblending(img1, img2, mask):
"""Helper: interface to external poisson blending"""
return blending.blend(img1, img2, 1 - mask)
@staticmethod
def iminvtransform(img):
"""Helper: Rescale pixel value ranges to 0 and 1"""
return (np.array(img) + 1.0) / 2.0
def loss_LSGAN(y_true, y_pred):
return K.mean(K.square(y_pred-y_true), axis=-1)/2