-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmodel.py
219 lines (183 loc) · 7.79 KB
/
model.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
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
# -*- coding: UTF-8 -*-
from utils import (
read_data,
input_setup,
imsave,
merge,
imread,
merge2,
psnr,
psnr2
)
import time
import os
import matplotlib.pyplot as plt
import numpy as np
import tensorflow as tf
import glob
import feature_map
import csv
import partial_conv as pc
try:
xrange
except:
xrange = range
class SRCNN(object):
def __init__(self,
sess,
image_size=33,
label_size=33,
batch_size=128,
c_dim=1,
checkpoint_dir=None,
sample_dir=None):
self.sess = sess
self.is_grayscale = (c_dim == 1)
self.image_size = image_size
self.label_size = label_size
self.batch_size = batch_size
self.c_dim = c_dim
self.checkpoint_dir = checkpoint_dir
self.sample_dir = sample_dir
self.build_model()
def build_model(self):
self.images = tf.placeholder(tf.float32, [None, self.image_size, self.image_size, self.c_dim], name='images')
self.labels = tf.placeholder(tf.float32, [None, self.label_size, self.label_size, self.c_dim], name='labels')
self.weights = {
'w1': tf.Variable(tf.random_normal([9, 9, 1, 64], stddev=1e-3, seed = 12345), name='w1'),
'w2': tf.Variable(tf.random_normal([1, 1, 64, 32], stddev=1e-3, seed = 54321), name='w2'),
'w3': tf.Variable(tf.random_normal([5, 5, 32, 1], stddev=1e-3, seed = 11111), name='w3')
}
self.biases = {
'b1': tf.Variable(tf.zeros([64]), name='b1'),
'b2': tf.Variable(tf.zeros([32]), name='b2'),
'b3': tf.Variable(tf.zeros([1]), name='b3')
}
self.pred = self.model()
# Loss function (MSE)
self.loss = tf.reduce_mean(tf.square(self.labels - self.pred))
# a = tf.zeros([17,17,self.c_dim],dtype=tf.dtypes.float32)
# self.labels2 = self.labels
# self.pred2 = self.pred
# self.labels2[-1,8:25,8:25,self.c_dim] = a
# self.pred2[-1,8:25,8:25,self.c_dim] = a
#
# self.loss2 = tf.reduce_mean(tf.square(self.labels2 - self.pred2))
self.saver = tf.train.Saver()
def train(self, config):
if config.is_train:
input_setup(self.sess, config)
else:
nx, ny, h, w, h_, w_, temp_h2, temp_w2 = input_setup(self.sess, config)
if config.is_train:
data_dir = os.path.join('./{}'.format(config.checkpoint_dir), "train.h5_test")
else:
data_dir = os.path.join('./{}'.format(config.checkpoint_dir), "test.h5_test")
train_data, train_label = read_data(data_dir)
# Stochastic gradient descent with the standard backpropagation
self.train_op = tf.train.GradientDescentOptimizer(config.learning_rate).minimize(self.loss)
# self.train_op2 = tf.train.GradientDescentOptimizer(config.learning_rate).minimize(self.loss2)
tf.initialize_all_variables().run()
counter = 0
start_time = time.time()
if self.load(self.checkpoint_dir):
print(" [*] Load SUCCESS")
else:
print(" [!] Load failed...")
if config.is_train:
print("Training...")
for ep in xrange(config.epoch):
# Run by batch images
batch_idxs = len(train_data) // config.batch_size
for idx in xrange(0, batch_idxs):
batch_images = train_data[idx*config.batch_size : (idx+1)*config.batch_size]
batch_labels = train_label[idx*config.batch_size : (idx+1)*config.batch_size]
counter += 1
_, err = self.sess.run([self.train_op, self.loss], feed_dict={self.images: batch_images, self.labels: batch_labels})
if counter % 10 == 0:
print("Epoch: [%2d], step: [%2d], time: [%4.4f], loss: [%.8f]" \
% ((ep+1), counter, time.time()-start_time, err))
loss_ = []
loss_.append(err)
if counter % 500 == 0:
self.save(config.checkpoint_dir, counter)
with open('loss_residual.csv','a',newline='') as csvfile:
writer = csv.writer(csvfile)
np.array(loss_)
writer.writerow([float(np.mean(loss_))])
else:
print("Testing...")
visualize_layers = ['conv1','conv2','conv3']
conv_out = self.sess.run(tf.get_collection('activations'), feed_dict={self.images: train_data, self.labels: train_label})
result = self.pred.eval({self.images: train_data, self.labels: train_label})
# print(merge(train_data,[nx,ny]).shape)
# print(merge(train_label,[nx,ny]).shape)
# print(len(result))
# print(np.shape(result))
# print(len(conv_out))
# print(type(conv_out))
# print(np.shape(conv_out[0]))
# print(np.shape(conv_out[1]))
# print(np.shape(conv_out[2]))
result = merge(result, [nx, ny])
result = result.squeeze()
result_ = result[temp_h2:h + temp_h2 ,temp_w2: w + temp_w2]
image_path = os.path.join(os.getcwd(), config.sample_dir)
image_path = os.path.join(image_path, "set5-15000-0.png")
imsave(result_, image_path)
test = merge(train_label, [nx, ny])
test_ = test[temp_h2:h + temp_h2 ,temp_w2: w + temp_w2]
print('PSNR : {}'.format(psnr(test_, result_)))
with open('psnr-set5.csv', 'a', newline='') as csvfile:
writer = csv.writer(csvfile)
writer.writerow([float(psnr(test_, result_))])
print('result_:{}'.format(np.shape(result_)))
print('test_:{}'.format(np.shape(test_)))
print(result_)
# feature = []
# feature2 = []
# feature3 = []
# feature_all = []
# for number_1 in range(len(conv_out[0][0,0,0,:])):
# feature.append((merge2(conv_out[0][:,:,:,number_1],[nx,ny])))
# for number_2 in range(len(conv_out[1][0,0,0,:])):
# feature2.append((merge2(conv_out[1][:,:,:,number_2],[nx,ny])))
# for number_3 in range(len(conv_out[2][0,0,0,:])):
# feature3.append((merge2(conv_out[2][:,:,:,number_3],[nx,ny])))
# feature_all.append(feature)
# feature_all.append(feature2)
# feature_all.append(feature3)
# for i, layer in enumerate(visualize_layers):
# plot_dir=r'C:\Users\GL63\Desktop\paper\SRCNN-Tensorflow-master-trial3\SRCNN-Tensorflow-master\feature_map_15000-'#要保存的路径
# if not os.path.exists(plot_dir+layer):#如果路径不存在，则创建文件夹
# os.mkdir(plot_dir+layer)
# for l in range(len(feature_all[i])):#保存为图片
# feature_map.plot_conv_output(feature_all[i], plot_dir + layer, str(l), filters_all=False, filters=[l])
def model(self):
conv1 = tf.nn.relu(pc.partial_conv(self.images, self.weights['w1'], strides=[1,1,1,1], padding='SAME') + self.biases['b1'])
conv2 = tf.nn.relu(pc.partial_conv(conv1, self.weights['w2'], strides=[1,1,1,1], padding='VALID') + self.biases['b2'])
conv3 = pc.partial_conv(conv2, self.weights['w3'], strides=[1,1,1,1], padding='SAME') + self.biases['b3']
tf.add_to_collection('activations', conv1)
tf.add_to_collection('activations', conv2)
tf.add_to_collection('activations', conv3)
return conv3
def save(self, checkpoint_dir, step):
model_name = "SRCNN.model"
model_dir = "%s_%s" % ("srcnn", self.label_size)
checkpoint_dir = os.path.join(checkpoint_dir, model_dir)
if not os.path.exists(checkpoint_dir):
os.makedirs(checkpoint_dir)
self.saver.save(self.sess,
os.path.join(checkpoint_dir, model_name),
global_step=step)
def load(self, checkpoint_dir):
print(" [*] Reading checkpoints...")
model_dir = "%s_%s" % ("srcnn", self.label_size)
checkpoint_dir = os.path.join(checkpoint_dir, model_dir)
ckpt = tf.train.get_checkpoint_state(checkpoint_dir)
if ckpt and ckpt.model_checkpoint_path:
ckpt_name = os.path.basename(ckpt.model_checkpoint_path)
self.saver.restore(self.sess, os.path.join(checkpoint_dir, ckpt_name))
return True
else:
return False