-
Notifications
You must be signed in to change notification settings - Fork 8
/
segnet.py
239 lines (184 loc) · 9.1 KB
/
segnet.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
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
from keras.models import Sequential
from keras import layers
from keras import models
from keras.layers import BatchNormalization, Conv2D, UpSampling2D, MaxPooling2D, Dropout
from keras.optimizers import Adam
from keras import regularizers
import numpy as np
import pickle
from keras import backend as K
from keras.layers import Layer
class MaxPoolingWithArgmax2D(Layer):
def __init__(self, pool_size=(2, 2), strides=(2, 2), padding='same', **kwargs):
super(MaxPoolingWithArgmax2D, self).__init__(**kwargs)
self.padding = padding
self.pool_size = pool_size
self.strides = strides
def call(self, inputs, **kwargs):
padding = self.padding
pool_size = self.pool_size
strides = self.strides
if K.backend() == 'tensorflow':
ksize = [1, pool_size[0], pool_size[1], 1]
padding = padding.upper()
strides = [1, strides[0], strides[1], 1]
output, argmax = K.tf.nn.max_pool_with_argmax(inputs, ksize=ksize, strides=strides, padding=padding)
else:
errmsg = '{} backend is not supported for layer {}'.format(K.backend(), type(self).__name__)
raise NotImplementedError(errmsg)
argmax = K.cast(argmax, K.floatx())
return [output, argmax]
def compute_output_shape(self, input_shape):
ratio = (1, 2, 2, 1)
output_shape = [dim // ratio[idx] if dim is not None else None for idx, dim in enumerate(input_shape)]
output_shape = tuple(output_shape)
return [output_shape, output_shape]
def compute_mask(self, inputs, mask=None):
return 2 * [None]
class MaxUnpooling2D(Layer):
def __init__(self, size=(2, 2), **kwargs):
super(MaxUnpooling2D, self).__init__(**kwargs)
self.size = size
def call(self, inputs, output_shape=None):
updates, mask = inputs[0], inputs[1]
with K.tf.variable_scope(self.name):
mask = K.cast(mask, 'int32')
input_shape = K.tf.shape(updates, out_type='int32')
# calculation new shape
if output_shape is None:
output_shape = (input_shape[0], input_shape[1] * self.size[0], input_shape[2] * self.size[1], input_shape[3])
self.output_shape1 = output_shape
# calculation indices for batch, height, width and feature maps
one_like_mask = K.ones_like(mask, dtype='int32')
batch_shape = K.concatenate([[input_shape[0]], [1], [1], [1]], axis=0)
batch_range = K.reshape(K.tf.range(output_shape[0], dtype='int32'), shape=batch_shape)
b = one_like_mask * batch_range
y = mask // (output_shape[2] * output_shape[3])
x = (mask // output_shape[3]) % output_shape[2]
feature_range = K.tf.range(output_shape[3], dtype='int32')
f = one_like_mask * feature_range
# transpose indices & reshape update values to one dimension
updates_size = K.tf.size(updates)
print "AAAAA", updates_size
indices = K.transpose(K.reshape(K.stack([b, y, x, f]), [4, updates_size]))
values = K.reshape(updates, [updates_size])
ret = K.tf.scatter_nd(indices, values, output_shape)
return ret
def compute_output_shape(self, input_shape):
mask_shape = input_shape[1]
return mask_shape[0], mask_shape[1] * self.size[0], mask_shape[2] * self.size[1], mask_shape[3]
def get_segnet():
concat_axis = 3
pool_size = (2,2)
inputs = layers.Input(shape = (80, 120, 3))
conv_1 = BatchNormalization()(inputs)
conv_1 = Conv2D(64, (3, 3), activation='relu', padding="same")(conv_1)
conv_1 = BatchNormalization()(conv_1)
conv_2 = Conv2D(64, (3, 3), activation='relu', padding="same")(conv_1)
conv_2 = BatchNormalization()(conv_2)
pool_1, mask_1 = MaxPoolingWithArgmax2D(pool_size)(conv_2)
conv_3 = Conv2D(128, (3, 3), activation='relu', padding="same")(pool_1)
conv_3 = BatchNormalization()(conv_3)
conv_4 = Conv2D(128, (3, 3), activation='relu', padding="same")(conv_3)
conv_4 = BatchNormalization()(conv_4)
pool_2, mask_2 = MaxPoolingWithArgmax2D(pool_size)(conv_4)
conv_5 = Conv2D(256, (3, 3), activation='relu', padding="same")(pool_2)
conv_5 = BatchNormalization()(conv_5)
conv_6 = Conv2D(256, (3, 3), activation='relu', padding="same")(conv_5)
conv_6 = BatchNormalization()(conv_6)
conv_7 = Conv2D(256, (3, 3), activation='relu', padding="same")(conv_6)
conv_7 = BatchNormalization()(conv_7)
pool_3, mask_3 = MaxPoolingWithArgmax2D(pool_size)(conv_7)
conv_8 = Conv2D(512, (3, 3), activation='relu', padding="same")(pool_3)
conv_8 = BatchNormalization()(conv_8)
conv_9 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_8)
conv_9 = BatchNormalization()(conv_9)
conv_10 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_9)
conv_10 = BatchNormalization()(conv_10)
#pool_4, mask_4 = MaxPoolingWithArgmax2D(pool_size)(conv_10)
pool_4, mask_4 = MaxPoolingWithArgmax2D(pool_size=(2, 3), strides=(2, 3))(conv_10)
conv_11 = Conv2D(512, (3, 3), activation='relu', padding="same")(pool_4)
conv_11 = BatchNormalization()(conv_11)
conv_12 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_11)
conv_12 = BatchNormalization()(conv_12)
conv_13 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_12)
conv_13 = BatchNormalization()(conv_13)
#pool_5, mask_5 = MaxPoolingWithArgmax2D(pool_size)(conv_13)
#print("Build enceder done..")
# decoder
#unpool_1 = MaxUnpooling2D(pool_size)([pool_5, mask_5])
conv_14 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_13)
conv_14 = BatchNormalization()(conv_14)
conv_15 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_14)
conv_15 = BatchNormalization()(conv_15)
conv_16 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_15)
conv_16 = BatchNormalization()(conv_16)
#unpool_2 = MaxUnpooling2D(pool_size)([conv_16, mask_4])
unpool_2 = MaxUnpooling2D((2,3))([conv_16, mask_4])
conv_17 = Conv2D(512, (3, 3), activation='relu', padding="same")(unpool_2)
#conv_17 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_10)
conv_17 = BatchNormalization()(conv_17)
conv_18 = Conv2D(512, (3, 3), activation='relu', padding="same")(conv_17)
conv_18 = BatchNormalization()(conv_18)
conv_19 = Conv2D(256, (3, 3), activation='relu', padding="same")(conv_18)
conv_19 = BatchNormalization()(conv_19)
unpool_3 = MaxUnpooling2D(pool_size)([conv_19, mask_3])
conv_20 = Conv2D(256, (3, 3), activation='relu', padding="same")(unpool_3)
conv_20 = BatchNormalization()(conv_20)
conv_21 = Conv2D(256, (3, 3), activation='relu', padding="same")(conv_20)
conv_21 = BatchNormalization()(conv_21)
conv_22 = Conv2D(128, (3, 3), activation='relu', padding="same")(conv_21)
conv_22 = BatchNormalization()(conv_22)
unpool_4 = MaxUnpooling2D(pool_size)([conv_22, mask_2])
conv_23 = Conv2D(128, (3, 3), activation='relu', padding="same")(unpool_4)
conv_23 = BatchNormalization()(conv_23)
conv_24 = Conv2D(64, (3, 3), activation='relu', padding="same")(conv_23)
conv_24 = BatchNormalization()(conv_24)
unpool_5 = MaxUnpooling2D(pool_size)([conv_24, mask_1])
conv_25 = Conv2D(64, (3, 3), activation='relu', padding="same")(unpool_5)
conv_25 = BatchNormalization()(conv_25)
conv_26 = Conv2D(1, (1, 1), activation='relu', padding="valid")(conv_25)
#conv_26 = Reshape((80, 120, 1), input_shape=(80, 120, 1))(conv_26)
#outputs = Activation(output_mode)(conv_26)
outputs = conv_26
print("Build decoder done..")
model = models.Model(inputs=inputs, outputs=outputs)
model.compile(loss='mae', optimizer=Adam(lr=0.001), metrics=['mse'])
print model.summary()
return model
#x = np.load("/datasets/1980-2016/z_1980_2016.npy")
x = np.load("/datasets/10zlevels.npy")
y = 1000*np.expand_dims(np.load("/datasets/1980-2016/full_tp_1980_2016.npy"), axis=3)
idxs = np.arange(x.shape[0])
np.random.seed(0)
np.random.shuffle(idxs)
x = x[idxs, :, :, :]
y = y[idxs, :]
y_train = y[:40000, :]
y_test = y[40000:, :]
# Levels [1000, 900, 800, 700, 600, 500, 400, 300, 200, 100]
i = 0
j = 2
k = 6
x_train = x[:40000, :, :, [i,j,k]]
x_test = x[40000:, :, :, [i,j,k]]
model = get_segnet()
history = model.fit(x_train, y_train, epochs=50, verbose=1, validation_data=(x_test, y_test))
with open('trainHistoryDict_{}-{}-{}'.format(i, j, k), 'wb') as file_hist:
pickle.dump(history.history, file_hist)
#model.save('/datasets/3chan_{}-{}-{}_.h5')
"""
for i in range(10):
rain_pred = model.predict(x_test[i:i+1,:])
rain = y_test[i:i+1,:]
print "Total pred", np.sum(rain_pred)
print "Total", np.sum(rain)
print "Mean pred", np.sum(rain_pred) / float(rain.shape[1] * rain.shape[2])
print "Mean", np.sum(rain) / float(rain.shape[1] * rain.shape[2])
mae = np.sum(rain_pred - rain)
print "E", mae
mae = np.sum(np.absolute(rain_pred - rain))
print "AE", mae
mae /= float(rain.shape[1] * rain.shape[2])
print "MAE", mae
"""