-
Notifications
You must be signed in to change notification settings - Fork 7
/
main_train.py
186 lines (145 loc) · 6.97 KB
/
main_train.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
"""
Version: v1.2
Date: 2021-01-12
Author: Mullissa A.G.
Description: This script trains a complex-valued multistream fully convolutional network for despeckling a
polarimetric SAR covariance matrix as discussed in
our paper A. G. Mullissa, C. Persello and J. Reiche,
"Despeckling Polarimetric SAR Data Using a Multistream Complex-Valued Fully Convolutional Network,"
in IEEE Geoscience and Remote Sensing Letters, doi: 10.1109/LGRS.2021.3066311.
Some utility functions are adopted from https://github.com/cszn/DnCNN
"""
# =============================================================================
import complexnn
import helper
import argparse
import re
import os, glob, datetime
import numpy as np
from keras.layers import Input, Add
from keras.models import Model, load_model
from keras.callbacks import CSVLogger, ModelCheckpoint, LearningRateScheduler
from keras.optimizers import Adam
import keras.backend as K
## Params
parser = argparse.ArgumentParser()
parser.add_argument('--model', default='cv-despecknet', type=str, help='choose a type of model')
parser.add_argument('--batch_size', default=128, type=int, help='batch size')
parser.add_argument('--train_image', default='data/Train2', type=str, help='path of train data real')
parser.add_argument('--train_label', default='data/Label2', type=str, help='path of label data')
parser.add_argument('--epoch', default=50, type=int, help='number of train epoches')
parser.add_argument('--lr', default=1e-3, type=float, help='initial learning rate for Adam')
parser.add_argument('--save_every', default=1, type=int, help='save model at every x epoches')
args = parser.parse_args()
save_dir = os.path.join('models',args.model)
if not os.path.exists(save_dir):
os.mkdir(save_dir)
def cv_deSpeckNet(depth,filters=48,image_channels=6, use_bnorm=True):
#FCN noise
layer_count = 0
inpt = Input(shape=(None,None,image_channels),name = 'input'+str(layer_count))
# 1st layer, CV-Conv+Crelu
layer_count += 1
x0 = complexnn.conv.ComplexConv2D(filters=filters, kernel_size=(3,3), strides=(1,1), activation='relu', padding='same',name = 'conv'+str(layer_count))(inpt)
# depth-2 layers, CV-Conv+CV-BN+Crelu
for i in range(depth-2):
layer_count += 1
x0 = complexnn.conv.ComplexConv2D(filters=filters, kernel_size=(3,3), strides=(1,1),activation='relu', padding='same',name = 'conv'+str(layer_count))(x0)
if use_bnorm:
layer_count += 1
x0 = complexnn.bn.ComplexBatchNormalization(name = 'bn'+str(layer_count))(x0)
# last layer, CV-Conv
layer_count += 1
x0 = complexnn.conv.ComplexConv2D(filters=3, kernel_size=(3,3), strides=(1,1),padding='same',name = 'speckle'+str(1))(x0)
layer_count += 1
#FCN clean
# 1st layer, CV-Conv+Crelu
x = complexnn.conv.ComplexConv2D(filters=filters, kernel_size=(3,3), strides=(1,1), activation='relu', padding='same',name = 'conv'+str(layer_count))(inpt)
# depth-2 layers, CV-Conv+CV-BN+Crelu
for i in range(depth-2):
layer_count += 1
x = complexnn.conv.ComplexConv2D(filters=filters, kernel_size=(3,3), strides=(1,1),activation='relu', padding='same',name = 'conv'+str(layer_count))(x)
if use_bnorm:
layer_count += 1
x = complexnn.bn.ComplexBatchNormalization(name = 'bn'+str(layer_count))(x)
# last layer, CV-Conv
layer_count += 1
x = complexnn.conv.ComplexConv2D(filters=3, kernel_size=(3,3), strides=(1,1),padding='same',name = 'clean'+str(1))(x)
layer_count += 1
x_orig = Add(name = 'noisy' + str(1))([x0,x])
model = Model(inputs=inpt, outputs=[x,x_orig])
return model
def findLastCheckpoint(save_dir):
file_list = glob.glob(os.path.join(save_dir,'model_*.hdf5')) # get name list of all .hdf5 files
#file_list = os.listdir(save_dir)
if file_list:
epochs_exist = []
for file_ in file_list:
result = re.findall(".*model_(.*).hdf5.*",file_)
#print(result[0])
epochs_exist.append(int(result[0]))
initial_epoch=max(epochs_exist)
else:
initial_epoch = 0
return initial_epoch
def log(args,kwargs):
print(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S:"),args,kwargs)
def lr_schedule(epoch):
initial_lr = args.lr
if epoch<=30:
lr = initial_lr
elif epoch<=60:
lr = initial_lr/10
elif epoch<=80:
lr = initial_lr/20
else:
lr = initial_lr/20
#log('current learning rate is %2.8f' %lr)
return lr
def train_datagen(epoch_iter=2000,epoch_num=5,batch_size=64,data_dir=args.train_image,label_dir=args.train_label):
while(True):
n_count = 0
if n_count == 0:
#print(n_count)
xs = helper.make_dataTensor(data_dir)
xy = helper.make_dataTensor(label_dir)
assert len(xs)%batch_size ==0, \
log('make sure the last iteration has a full batchsize, this is important if you use batch normalization!')
xs = xs.astype('float32')
xy = xy.astype('float32')
indices = list(range(xs.shape[0]))
n_count = 1
for _ in range(epoch_num):
np.random.shuffle(indices) # shuffle
for i in range(0, len(indices), batch_size):
batch_x = xs[indices[i:i+batch_size]]
batch_y = xy[indices[i:i+batch_size]]
yield batch_x, [batch_y, batch_x]
# sum square error loss function
def sum_squared_error(y_true, y_pred):
return K.sum(K.square(y_pred - y_true))/2
if __name__ == '__main__':
# model selection
model = cv_deSpeckNet(depth=17,filters=48,image_channels=6,use_bnorm=True)
model.summary()
# load the last model in matconvnet style
initial_epoch = findLastCheckpoint(save_dir=save_dir)
if initial_epoch > 0:
print('resuming by loading epoch %03d'%initial_epoch)
model = load_model(os.path.join(save_dir,'model_%03d.hdf5'%initial_epoch), custom_objects={'ComplexConv2D': complexnn.conv.ComplexConv2D, 'ComplexBatchNormalization': complexnn.bn.ComplexBatchNormalization, 'sum_squared_error': sum_squared_error})
loss_funcs = {
'clean1': sum_squared_error,
'noisy1' : sum_squared_error}
loss_weights = {'clean1': 100.0, 'noisy1': 1.0}
# compile the model
model.compile(optimizer=Adam(0.001), loss=loss_funcs, loss_weights=loss_weights)
# use call back functions
checkpointer = ModelCheckpoint(os.path.join(save_dir,'model_{epoch:03d}.hdf5'),
verbose=1, save_weights_only=False, period=1)
csv_logger = CSVLogger(os.path.join(save_dir,'log.csv'), append=True, separator=',')
lr_scheduler = LearningRateScheduler(lr_schedule)
# numer of steps per epoch
nsteps = helper.get_steps(args.train_image, batch_size=64)
history = model.fit_generator(train_datagen(batch_size=64),
steps_per_epoch=nsteps, epochs=51, verbose=1, initial_epoch=initial_epoch,
callbacks=[checkpointer,csv_logger,lr_scheduler])