def get_unet(n_ch,patch_height,patch_width):
inputs = Input(shape=(n_ch,patch_height,patch_width))
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv1)
pool1 = MaxPooling2D((2, 2))(conv1)
#
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv2)
pool2 = MaxPooling2D((2, 2))(conv2)
#
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same',data_format='channels_first')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv3)
up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = concatenate([conv2,up1],axis=1)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv4)
#
up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = concatenate([conv1,up2], axis=1)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv5)
#
conv6 = Conv2D(2, (1, 1), activation='relu',padding='same',data_format='channels_first')(conv5)
conv6 = core.Reshape((2,patch_height*patch_width))(conv6)
conv6 = core.Permute((2,1))(conv6)
############
conv7 = core.Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=conv7)
# sgd = SGD(lr=0.01, decay=1e-6, momentum=0.3, nesterov=False)
#if optimizer= 'sgd', the default values for the hyperparameters will be used.
model.compile(optimizer='sgd', loss='categorical_crossentropy',metrics=['accuracy'])
return model
def get_unet(n_ch,patch_height,patch_width):
inputs = Input(shape=(n_ch,patch_height,patch_width))
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv1)
pool1 = MaxPooling2D((2, 2))(conv1)
#
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv2)
pool2 = MaxPooling2D((2, 2))(conv2)
#
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same',data_format='channels_first')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv3)
up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = concatenate([conv2,up1],axis=1)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv4)
#
up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = concatenate([conv1,up2], axis=1)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv5)
#
conv6 = Conv2D(3, (1, 1), activation='relu',padding='same',data_format='channels_first')(conv5)
conv6 = core.Reshape((3,patch_height*patch_width))(conv6)
conv6 = core.Permute((2,1))(conv6)
############
conv7 = core.Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=conv7)
# sgd = SGD(lr=0.01, decay=1e-6, momentum=0.3, nesterov=False)
adam = Adam(lr=0.001, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=False)
model.compile(optimizer=adam,loss='categorical_crossentropy',metrics=['categorical_accuracy'],sample_weight_mode="temporal")
return model
def get_unet(n_ch,patch_height,patch_width):
inputs = Input((n_ch,patch_height, patch_width))
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(inputs)#'valid'
conv1 = Dropout(0.3)(conv1)
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
#
conv2 = Conv2D(64, (3, 3), padding='same')(pool1) #,activation='relu', padding='same')(pool1)
conv2 = normalization.BatchNormalization(epsilon=2e-05, axis=1, momentum=0.9, weights=None, beta_initializer='zero', gamma_initializer='one')(conv2)
conv2 = Activation('relu')(conv2)
#conv2 = Dropout(0.3)(conv2)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv2)#,W_regularizer=l2(0.01), b_regularizer=l2(0.01))(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
#
conv3 = Conv2D(128, (3, 3), padding='same')(pool2) #, activation='relu', padding='same')(pool2)
conv3 = normalization.BatchNormalization(epsilon=2e-05,axis=1, momentum=0.9, weights=None, beta_initializer='zero', gamma_initializer='one')(conv3)
conv3 = Activation('relu')(conv3)
#conv3 = Dropout(0.3)(conv3)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(conv3)#,W_regularizer=l2(0.01), b_regularizer=l2(0.01))(conv3)
up1 = concatenate([UpSampling2D(size=(2, 2))(conv3), conv2], axis=1)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same')(up1)
conv4 = Dropout(0.3)(conv4)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv4)
#
up2 = concatenate([UpSampling2D(size=(2, 2))(conv4), conv1], axis=1)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same')(up2)
conv5 = Dropout(0.3)(conv5)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv5)
conv6 = Conv2D(2, (1, 1), activation='relu',padding='same')(conv5)
#conv6 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(conv6)
conv6 = core.Reshape((2,patch_height*patch_width))(conv6)
conv6 = core.Permute((2,1))(conv6)
############
act = Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=act)
return model
def get_unet(n_ch,patch_height,patch_width):
inputs = Input(shape=(n_ch,patch_height,patch_width))
conv1 = Conv2D(32, 3, 3, activation='relu', border_mode='same')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, 3, 3, activation='relu', border_mode='same')(conv1)
pool1 = MaxPooling2D((2, 2))(conv1)
#
conv2 = Conv2D(64, 3, 3, activation='relu', border_mode='same')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, 3, 3, activation='relu', border_mode='same')(conv2)
pool2 = MaxPooling2D((2, 2))(conv2)
#
conv3 = Conv2D(128, 3, 3, activation='relu', border_mode='same')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, 3, 3, activation='relu', border_mode='same')(conv3)
up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = merge([conv2,up1], mode='concat', concat_axis=1)
conv4 = Conv2D(64, 3, 3, activation='relu', border_mode='same')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, 3, 3, activation='relu', border_mode='same')(conv4)
#
up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = merge([conv1,up2], mode='concat', concat_axis=1)
conv5 = Conv2D(32, 3, 3, activation='relu', border_mode='same')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, 3, 3, activation='relu', border_mode='same')(conv5)
#
conv6 = Conv2D(2, 1, 1, activation='relu', border_mode='same')(conv5)
conv6 = core.Reshape((2,patch_height*patch_width))(conv6)
conv6 = core.Permute((2,1))(conv6)
############
conv7 = core.Activation('softmax')(conv6)
model = Model(input=inputs, output=conv7)
# sgd = SGD(lr=0.01, decay=1e-6, momentum=0.3, nesterov=False)
#model.compile(optimizer='sgd', loss='categorical_crossentropy',metrics=['accuracy'], context=['gpu(0)','gpu(1)','gpu(2)'])
model.compile(optimizer='sgd', loss='categorical_crossentropy',metrics=['accuracy'], context=['gpu(0)'])
return model
def get_unet(n_ch,patch_height,patch_width):
inputs = Input(shape=(n_ch,patch_height,patch_width))
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv1)
pool1 = MaxPooling2D((2, 2))(conv1)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv2)
pool2 = MaxPooling2D((2, 2))(conv2)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same',data_format='channels_first')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv3)
up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = concatenate([conv2,up1],axis=1)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv4)
up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = concatenate([conv1,up2], axis=1)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same',data_format='channels_first')(conv5)
conv6 = Conv2D(2, (1, 1), activation='relu',padding='same',data_format='channels_first')(conv5)
conv6 = core.Reshape((2,patch_height*patch_width))(conv6)
conv6 = core.Permute((2,1))(conv6)
conv7 = core.Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=conv7)
model.compile(optimizer='sgd', loss='categorical_crossentropy',metrics=['accuracy'])
return model
def Unet (nClasses , optimizer=None , input_width=360 , input_height=480 , nChannels=1 ):
inputs = Input((nChannels, input_height, input_width))
conv1 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
conv2 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
conv3 = Convolution2D(128, 3, 3, activation='relu', border_mode='same')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Convolution2D(128, 3, 3, activation='relu', border_mode='same')(conv3)
up1 = merge([UpSampling2D(size=(2, 2))(conv3), conv2], mode='concat', concat_axis=1)
conv4 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(conv4)
up2 = merge([UpSampling2D(size=(2, 2))(conv4), conv1], mode='concat', concat_axis=1)
conv5 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv5)
conv6 = Convolution2D(nClasses, 1, 1, activation='relu',border_mode='same')(conv5)
conv6 = core.Reshape((nClasses,input_height*input_width))(conv6)
conv6 = core.Permute((2,1))(conv6)
conv7 = core.Activation('softmax')(conv6)
model = Model(input=inputs, output=conv7)
if not optimizer is None:
model.compile(loss="categorical_crossentropy", optimizer= optimizer , metrics=['accuracy'] )
return model
def Unet(nClasses, optimizer=None, input_width=64, input_height=96, nChannels=1):
inputs = Input((input_height, input_width, nChannels))
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(conv3)
up1 = concatenate([UpSampling2D(size=(2, 2))(conv3), conv2], axis=-1)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv4)
up2 = concatenate([UpSampling2D(size=(2, 2))(conv4), conv1], axis=-1)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv5)
conv6 = Conv2D(nClasses, (1, 1), activation='relu', padding='same')(conv5)
conv6 = core.Reshape((nClasses, input_height * input_width))(conv6)
conv6 = core.Permute((2, 1))(conv6)
conv7 = core.Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=conv7)
if not optimizer is None:
model.compile(loss="categorical_crossentropy", optimizer=optimizer, metrics=['accuracy'])
return model
def set_cnn_model_attention(input_dim=4, input_length=2701):
attention_reg_x = 0.25
attention_reg_xr = 1
attentionhidden_x = 16
attentionhidden_xr = 8
nbfilter = 16
input = Input(shape=(input_length, input_dim))
x = conv.Convolution1D(nbfilter, 10, border_mode="valid")(input)
x = Dropout(0.5)(x)
x = Activation('relu')(x)
x = conv.MaxPooling1D(pool_length=3)(x)
x_reshape = core.Reshape((x._keras_shape[2], x._keras_shape[1]))(x)
x = Dropout(0.5)(x)
x_reshape = Dropout(0.5)(x_reshape)
decoder_x = Attention(hidden=attentionhidden_x,
activation='linear') # success
decoded_x = decoder_x(x)
output_x = myFlatten(x._keras_shape[2])(decoded_x)
decoder_xr = Attention(hidden=attentionhidden_xr, activation='linear')
decoded_xr = decoder_xr(x_reshape)
output_xr = myFlatten(x_reshape._keras_shape[2])(decoded_xr)
output = merge([output_x, output_xr, Flatten()(x)], mode='concat')
#output = BatchNormalization()(output)
output = Dropout(0.5)(output)
print output.shape
output = Dense(nbfilter * 10, activation="relu")(output)
output = Dropout(0.5)(output)
out = Dense(2, activation='softmax')(output)
#output = BatchNormalization()(output)
model = Model(input, out)
model.compile(loss='categorical_crossentropy', optimizer='rmsprop')
return model
def create_nn_architectire(n_ch, patch_height, patch_width):
inputs = Input((n_ch, patch_height, patch_width))
conv1 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
#
conv2 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
#
conv3 = Convolution2D(128, 3, 3, activation='relu', border_mode='same')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Convolution2D(128, 3, 3, activation='relu', border_mode='same')(conv3)
up1 = merge([UpSampling2D(size=(2, 2))(conv3), conv2], mode='concat', concat_axis=1)
conv4 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(conv4)
#
up2 = merge([UpSampling2D(size=(2, 2))(conv4), conv1], mode='concat', concat_axis=1)
conv5 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv5)
#
conv6 = Convolution2D(2, 1, 1, activation='relu',border_mode='same')(conv5)
conv6 = core.Reshape((2,patch_height*patch_width))(conv6)
conv6 = core.Permute((2,1))(conv6)
############
conv7 = core.Activation('softmax')(conv6)
model = Model(input=inputs, output=conv7)
model.compile(optimizer='sgd', loss='categorical_crossentropy',metrics=['accuracy'])
return model
def get_unet(n_ch, patch_height, patch_width):
inputs = Input((patch_height, patch_width, n_ch))
conv1 = Conv2DReluBatchNorm(32, 3, 3, inputs)
conv1 = Conv2DReluBatchNorm(32, 3, 3, conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
conv2 = Conv2DReluBatchNorm(64, 3, 3, pool1)
conv2 = Conv2DReluBatchNorm(64, 3, 3, conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
conv3 = Conv2DReluBatchNorm(128, 3, 3, pool2)
conv3 = Conv2DReluBatchNorm(128, 3, 3, conv3)
pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)
conv4 = Conv2DReluBatchNorm(256, 3, 3, pool3)
conv4 = Conv2DReluBatchNorm(256, 3, 3, conv4)
pool4 = MaxPooling2D(pool_size=(2, 2))(conv4)
conv5 = Conv2DReluBatchNorm(512, 3, 3, pool4)
conv5 = Conv2DReluBatchNorm(512, 3, 3, conv5)
up6 = merge([UpSampling2D(size=(2, 2))(conv5), conv4],
mode='concat',
concat_axis=3)
conv6 = Conv2DReluBatchNorm(256, 3, 3, up6)
conv6 = Conv2DReluBatchNorm(256, 3, 3, conv6)
up7 = merge([UpSampling2D(size=(2, 2))(conv6), conv3],
mode='concat',
concat_axis=3)
conv7 = Conv2DReluBatchNorm(128, 3, 3, up7)
conv7 = Conv2DReluBatchNorm(128, 3, 3, conv7)
up8 = merge([UpSampling2D(size=(2, 2))(conv7), conv2],
mode='concat',
concat_axis=3)
conv8 = Conv2DReluBatchNorm(64, 3, 3, up8)
conv8 = Conv2DReluBatchNorm(64, 3, 3, conv8)
up9 = merge([UpSampling2D(size=(2, 2))(conv8), conv1],
mode='concat',
concat_axis=3)
conv9 = Conv2DReluBatchNorm(32, 3, 3, up9)
conv9 = Conv2DReluBatchNorm(32, 3, 3, conv9)
conv10 = Convolution2D(1,
1,
1,
activation='sigmoid',
W_regularizer=l2(0.01))(conv9)
reshaped = core.Reshape((patch_height, patch_width))(conv10)
# conv10 = core.Permute((2, 1))(conv10)
model = Model(input=inputs, output=reshaped)
# model.compile(optimizer=Adam(lr=1e-5), loss=dice_coef_loss, metrics=['accuracy', jaccard_coef])
# model.compile(optimizer=Adam(lr=1e-5), loss='binary_crossentropy',
# metrics=['accuracy', jaccard_coef])
# model.compile(optimizer=Adam(lr=1e-5), loss=jaccard_loss, metrics=['accuracy', jaccard_coef])
model.compile(optimizer=Adam(lr=1e-5),
loss=binary_crossentropy_weight,
metrics=['accuracy', jaccard_coef])
return model
def test_reshape(self):
layer = core.Reshape(dims=(10, 10))
self._runner(layer)
def get_unet(n_ch, patch_height, patch_width):
inputs = Input(shape=(n_ch, patch_height, patch_width))
#data_format：字符串，“channels_first”或“channels_last”之一，代表图像的通道维的位置。
#以128x128的RGB图像为例，“channels_first”应将数据组织为（3,128,128），而“channels_last”应将数据组织为（128,128,3）。该参数的默认值是~/.keras/keras.json中设置的值，若从未设置过，则为“channels_last”。
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv1)
pool1 = MaxPooling2D((2, 2))(conv1)
#
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv2)
pool2 = MaxPooling2D((2, 2))(conv2)
#
conv3 = Conv2D(128, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv3)
up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = concatenate([conv2, up1], axis=1)
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv4)
#
up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = concatenate([conv1, up2], axis=1)
conv5 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv5)
#
#1×1的卷积的作用
#大概有两个方面的作用：1. 实现跨通道的交互和信息整合2. 进行卷积核通道数的降维和升维。
conv6 = Conv2D(2, (1, 1),
activation='relu',
padding='same',
data_format='channels_first')(conv5)
conv6 = core.Reshape((2, patch_height * patch_width))(
conv6) #此时output的shape是(batchsize,2,patch_height*patch_width)
conv6 = core.Permute((2, 1))(
conv6
) #此时output的shape是(Npatch,patch_height*patch_width,2)即输出维度是(Npatch,2304,2)
############
conv7 = core.Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=conv7)
# sgd = SGD(lr=0.01, decay=1e-6, momentum=0.3, nesterov=False)
model.compile(optimizer=Adam(lr=0.001),
loss='categorical_crossentropy',
metrics=['accuracy'])
return model
'''
def DeepCleave_CNN(output_model_name,
trainX,
trainY,
valX=None,
valY=None,
batch_size=1024,
epochs=500,
n_earlystop=None,
n_transfer_layer=1,
background_weights=None,
for_transfer=False,
compiletimes=0,
compilemodels=None,
predict=False):
input_row = trainX.shape[1]
input_col = trainX.shape[2]
trainX_t = trainX
valX_t = valX
checkpoint = ModelCheckpoint(filepath=output_model_name,
monitor='val_acc',
verbose=1,
mode='max',
save_best_only='True')
if (n_earlystop is not None):
early_stopping = EarlyStopping(monitor='val_acc', patience=n_earlystop)
epochs = 10000
#set to a very big value since earlystop used
callback_lists = [early_stopping, checkpoint]
else:
callback_lists = [checkpoint]
trainX_t.shape = (trainX_t.shape[0], input_row, input_col)
if (valX is not None):
valX_t.shape = (valX_t.shape[0], input_row, input_col)
if compiletimes == 0:
filtersize1 = 1
filtersize2 = 9
filtersize3 = 10
filter1 = 200
filter2 = 150
filter3 = 200
dropout1 = 0.75
dropout2 = 0.75
dropout4 = 0.75
dropout5 = 0.75
dropout6 = 0.25
L1CNN = 0
l2value = 0.001
nb_classes = 2
actfun = "relu"
optimization = 'adam'
attentionhidden_x = 10
attentionhidden_xr = 8
attention_reg_x = 0.151948
attention_reg_xr = 2
dense_size1 = 149
dense_size2 = 8
dropout_dense1 = 0.298224
dropout_dense2 = 0
input2 = Input(shape=(input_row, input_col))
x = conv.Convolution1D(filter1,
filtersize1,
init='he_normal',
W_regularizer=l2(l2value),
border_mode="same")(input2)
x = Dropout(dropout1)(x)
x1 = Activation(actfun)(x)
y1 = conv.Convolution1D(filter2,
filtersize2,
init='he_normal',
W_regularizer=l2(l2value),
border_mode="same")(x1)
y1 = Dropout(dropout2)(y1)
y1 = Activation(actfun)(y1)
y2 = conv.Convolution1D(filter2,
6,
init='he_normal',
W_regularizer=l2(l2value),
border_mode="same")(x1)
y2 = Dropout(dropout2)(y2)
y2 = Activation(actfun)(y2)
y3 = conv.Convolution1D(filter2,
3,
init='he_normal',
W_regularizer=l2(l2value),
border_mode="same")(x1)
y3 = Dropout(dropout2)(y3)
y3 = Activation(actfun)(y3)
mergeY = merge([y1, y2, y3], mode='concat', concat_axis=-1)
mergeY = Dropout(0.75)(mergeY)
z1 = conv.Convolution1D(filter3,
filtersize3,
init='he_normal',
W_regularizer=l2(l2value),
border_mode="same")(mergeY)
z1 = Activation(actfun)(z1)
z2 = conv.Convolution1D(filter3,
5,
init='he_normal',
W_regularizer=l2(l2value),
border_mode="same")(mergeY)
z2 = Activation(actfun)(z2)
z3 = conv.Convolution1D(filter3,
15,
init='he_normal',
W_regularizer=l2(l2value),
border_mode="same")(mergeY)
z3 = Activation(actfun)(z3)
x_reshape = core.Reshape((z1._keras_shape[2], z1._keras_shape[1]))(z1)
x = Dropout(dropout4)(z1)
x_reshape = Dropout(dropout5)(x_reshape)
decoder_x = Attention(hidden=attentionhidden_x,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_x)) # success
decoded_x = decoder_x(x)
output_x = myFlatten(x._keras_shape[2])(decoded_x)
decoder_xr = Attention(hidden=attentionhidden_xr,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_xr))
decoded_xr = decoder_xr(x_reshape)
output_xr = myFlatten(x_reshape._keras_shape[2])(decoded_xr)
x_reshape = core.Reshape((z2._keras_shape[2], z2._keras_shape[1]))(z2)
x = Dropout(dropout4)(z2)
x_reshape = Dropout(dropout5)(x_reshape)
decoder_x = Attention(hidden=attentionhidden_x,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_x)) # success
decoded_x = decoder_x(x)
output_x2 = myFlatten(x._keras_shape[2])(decoded_x)
decoder_xr = Attention(hidden=attentionhidden_xr,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_xr))
decoded_xr = decoder_xr(x_reshape)
output_xr2 = myFlatten(x_reshape._keras_shape[2])(decoded_xr)
x_reshape = core.Reshape((z3._keras_shape[2], z3._keras_shape[1]))(z3)
x = Dropout(dropout4)(z3)
x_reshape = Dropout(dropout5)(x_reshape)
decoder_x = Attention(hidden=attentionhidden_x,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_x)) # success
decoded_x = decoder_x(x)
output_x3 = myFlatten(x._keras_shape[2])(decoded_x)
decoder_xr = Attention(hidden=attentionhidden_xr,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_xr))
decoded_xr = decoder_xr(x_reshape)
output_xr3 = myFlatten(x_reshape._keras_shape[2])(decoded_xr)
output = merge([
output_x, output_xr, output_x2, output_xr2, output_x3, output_xr3
],
mode='concat')
output = Dropout(dropout6)(output)
output = Dense(dense_size1, init='he_normal',
activation='relu')(output)
output = Dropout(dropout_dense1)(output)
output = Dense(dense_size2, activation="relu",
init='he_normal')(output)
output = Dropout(dropout_dense2)(output)
out = Dense(nb_classes, init='he_normal', activation='softmax')(output)
cnn = Model(input2, out)
cnn.compile(loss='binary_crossentropy',
optimizer=optimization,
metrics=['accuracy'])
else:
cnn = compilemodels
if (predict is False):
if (background_weights is not None
and compiletimes == 0): #for the first time
if not for_transfer:
cnn.load_weights(background_weights)
else:
cnn2 = Model(input2, out)
cnn2.compile(loss='binary_crossentropy',
optimizer=optimization,
metrics=['accuracy'])
cnn2.load_weights(background_weights)
for l in range(
(len(cnn2.layers) -
n_transfer_layer)): #the last cnn is not included
cnn.layers[l].set_weights(cnn2.layers[l].get_weights())
cnn.layers[l].trainable = False # for frozen layer
cnn.compile(loss='binary_crossentropy',
optimizer=optimization,
metrics=['accuracy'])
if (valX is not None):
if (n_earlystop is None):
fitHistory = cnn.fit(trainX_t,
trainY,
batch_size=batch_size,
epochs=epochs,
validation_data=(valX_t, valY))
else:
fitHistory = cnn.fit(trainX_t,
trainY,
batch_size=batch_size,
epochs=epochs,
validation_data=(valX_t, valY),
callbacks=callback_lists)
else:
fitHistory = cnn.fit(trainX_t,
trainY,
batch_size=batch_size,
epochs=epochs)
return cnn
def build_model(self):
inputs = Input((self.patch_height, self.patch_width, 1))
conv1 = Conv2D(32, (3, 3), padding='same')(inputs) # 'valid'
conv1 = LeakyReLU(alpha=0.3)(conv1)
conv1 = Dropout(0.2)(conv1)
conv1 = normalization.BatchNormalization(
epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv1)
conv1 = Conv2D(32, (3, 3), dilation_rate=2, padding='same')(conv1)
conv1 = LeakyReLU(alpha=0.3)(conv1)
conv1 = Conv2D(32, (3, 3), dilation_rate=4, padding='same')(conv1)
conv1 = LeakyReLU(alpha=0.3)(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
# pool1 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(pool1)
conv2 = Conv2D(64, (3, 3), padding='same')(
pool1) # ,activation='relu', padding='same')(pool1)
conv2 = normalization.BatchNormalization(
epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv2)
conv2 = LeakyReLU(alpha=0.3)(conv2)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3), dilation_rate=2, padding='same')(conv2)
conv2 = LeakyReLU(alpha=0.3)(conv2)
conv2 = Conv2D(64, (3, 3), dilation_rate=4, padding='same')(
conv2) # ,W_regularizer=l2(0.01), b_regularizer=l2(0.01))(conv2)
conv2 = LeakyReLU(alpha=0.3)(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
# crop = Cropping2D(cropping=((int(3 * patch_height / 8), int(3 * patch_height / 8)), (int(3 * patch_width / 8), int(3 * patch_width / 8))))(conv1)
# conv3 = concatenate([crop,pool2], axis=1)
conv3 = Conv2D(128, (3, 3), padding='same')(
pool2) # , activation='relu', padding='same')(conv3)
conv3 = normalization.BatchNormalization(
epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv3)
conv3 = LeakyReLU(alpha=0.3)(conv3)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3), dilation_rate=2, padding='same')(
conv3) # ,W_regularizer=l2(0.01), b_regularizer=l2(0.01))(conv3)
conv3 = normalization.BatchNormalization(
epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv3)
conv3 = LeakyReLU(alpha=0.3)(conv3)
conv3 = Conv2D(128, (3, 3), dilation_rate=4, padding='same')(conv3)
conv3 = normalization.BatchNormalization(
epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv3)
conv3 = LeakyReLU(alpha=0.3)(conv3)
# up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = concatenate([UpSampling2D(size=(2, 2))(conv3), conv2], axis=3)
conv4 = Conv2D(64, (3, 3), padding='same')(up1)
conv4 = LeakyReLU(alpha=0.3)(conv4)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3), padding='same')(conv4)
conv4 = LeakyReLU(alpha=0.3)(conv4)
# conv4 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(conv4)
#
# up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = concatenate([UpSampling2D(size=(2, 2))(conv4), conv1], axis=3)
conv5 = Conv2D(32, (3, 3), padding='same')(up2)
conv5 = LeakyReLU(alpha=0.3)(conv5)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3), padding='same')(conv5)
conv5 = LeakyReLU(alpha=0.3)(conv5)
conv6 = Conv2D(self.num_seg_class + 1, (1, 1), padding='same')(conv5)
conv6 = LeakyReLU(alpha=0.3)(conv6)
# conv6 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(conv6)
# for tensorflow
# conv6 = core.Reshape((patch_height*patch_width,num_lesion_class+1))(conv6)
# for theano
conv6 = core.Reshape((self.patch_height * self.patch_width,
self.num_seg_class + 1))(conv6)
#conv6 = core.Permute((2, 1))(conv6)
############
act = Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=act)
model.compile(optimizer='adam',
loss='categorical_crossentropy',
metrics=['categorical_accuracy'])
# self.config.checkpoint = "C:\\Users\\kk\\Desktop\\Optic-Disc-Unet-master\\Optic-Disc-Unet-master\\experiments\\OpticDisc\\checkpoint"
plot_model(model,
to_file=os.path.join(self.config.checkpoint, "model.png"),
show_shapes=True)
self.model = model
# base_url = "" # linux
ratings = pd.read_csv(base_url + "ml-1m/ratings.dat",
sep='::',
names=['user_id', 'movie_id', 'rating', 'timestamp'])
n_users = np.max(ratings['user_id'])
n_movies = np.max(ratings['movie_id'])
print([n_users, n_movies, len(ratings)]) # ['userId', 'movieId', 100005]
plt.hist(ratings['rating'])
plt.show()
print(np.mean(ratings['rating']))
# 第一个小网络，处理用户嵌入层
model1 = Sequential()
model1.add(Embedding(n_users + 1, k, input_length=1))
model1.add(core.Reshape((k, ))) # keras.layers.core.Reshape
# 第二个小网络，处理电影嵌入层
model2 = Sequential()
model2.add(Embedding(n_movies + 1, k, input_length=1))
model2.add(core.Reshape((k, )))
# 第三个小网络，在第一，二个网络基础上叠加乘积运算
model = Sequential()
model.add(merge([model1, model2], mode='dot', dot_axes=1))
# 输出层和最后评分作对比，后向传播更新网络参数
model.compile(loss='mse', optimizer='adam')
# 另外可以尝试 optimizer = 'rmsprop' 或 'adagrad'
# 获取用户索引数据和电影索引数据，相应的特征矩阵X_train需要两个索引数据一起构造
def get_unet2(n_ch, patch_height, patch_width):
inputs = Input(shape=(n_ch, patch_height, patch_width))
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(inputs)
conv1 = BatchNormalization(axis=1)(conv1)
conv1 = Dropout(0.3)(conv1)
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv1)
conv1 = BatchNormalization(axis=1)(conv1)
pool1 = MaxPooling2D((2, 2), data_format='channels_first')(conv1)
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(pool1)
conv2 = BatchNormalization(axis=1)(conv2)
conv2 = Dropout(0.3)(conv2)
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv2)
conv2 = BatchNormalization(axis=1)(conv2)
pool2 = MaxPooling2D((2, 2), data_format='channels_first')(conv2)
conv3 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(pool2)
conv3 = BatchNormalization(axis=1)(conv3)
conv3 = Dropout(0.3)(conv3)
conv3 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv3)
conv3 = BatchNormalization(axis=1)(conv3)
pool3 = MaxPooling2D((2, 2), data_format='channels_first')(conv3)
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(pool3)
conv4 = BatchNormalization(axis=1)(conv4)
conv4 = Dropout(0.3)(conv4)
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv4)
conv4 = BatchNormalization(axis=1)(conv4)
pool4 = MaxPooling2D((2, 2), data_format='channels_first')(conv4)
conv5 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(pool4)
conv5 = BatchNormalization(axis=1)(conv5)
conv5 = Dropout(0.3)(conv5)
conv5 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv5)
conv5 = BatchNormalization(axis=1)(conv5)
up1 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv5)
up1 = concatenate([conv4, up1], axis=1)
conv6 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up1)
conv6 = BatchNormalization(axis=1)(conv6)
conv6 = Dropout(0.3)(conv6)
conv6 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv6)
conv6 = BatchNormalization(axis=1)(conv6)
up2 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv6)
up2 = concatenate([conv3, up2], axis=1)
conv7 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up2)
conv7 = BatchNormalization(axis=1)(conv7)
conv7 = Dropout(0.3)(conv7)
conv7 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv7)
conv7 = BatchNormalization(axis=1)(conv7)
up3 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv7)
up3 = concatenate([conv2, up3], axis=1)
conv8 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up3)
conv8 = BatchNormalization(axis=1)(conv8)
conv8 = Dropout(0.3)(conv8)
conv8 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv8)
conv8 = BatchNormalization(axis=1)(conv8)
up4 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv8)
up4 = concatenate([conv1, up4], axis=1)
conv9 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up4)
conv9 = BatchNormalization(axis=1)(conv9)
conv9 = Dropout(0.3)(conv9)
conv9 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv9)
conv9 = BatchNormalization(axis=1)(conv9)
conv10 = Conv2D(2, (1, 1),
activation='relu',
padding='same',
data_format='channels_first')(conv9)
conv10 = BatchNormalization(axis=1)(conv10)
conv10 = core.Reshape((2, patch_height * patch_width))(conv10)
conv10 = core.Permute((2, 1))(conv10)
############
conv10 = core.Activation('softmax')(conv10)
model = Model(input=inputs, output=conv10)
adaGrad = Adagrad(lr=1e-7, epsilon=1e-7, decay=1e-6)
model.compile(optimizer='sgd',
loss='categorical_crossentropy',
metrics=['accuracy'])
return model
up3 = concatenate([conv0, up3], axis=1)
conv6 = Conv2D(16, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up3)
conv6 = Dropout(0.2)(conv6)
conv6 = Conv2D(16, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv6)
conv7 = Conv2D(3, (1, 1),
activation='relu',
padding='same',
data_format='channels_first')(conv6)
conv7 = core.Reshape((3, 128 * 128))(conv7)
#conv6 = core.Permute((3,1))(conv6)
conv7 = core.Flatten()(conv7)
#conv7 = core.Dense(64)(conv7)
#conv7 = core.Activation('relu')(conv7)
#conv7 = Dropout(0.2)(conv7)
conv7 = core.Dense(2)(conv7)
############
conv8 = core.Activation('softmax')(conv7)
model = Model(input=inputs, output=conv8)
# sgd = SGD(lr=0.01, decay=1e-6, momentum=0.3, nesterov=False)
model.compile(optimizer='adam',
loss='categorical_crossentropy',
def get_unet(n_ch, patch_height, patch_width):
inputs = Input((n_ch, patch_height, patch_width))
conv1 = Conv2D(32, (3, 3), padding='same')(inputs) #'valid'
conv1 = LeakyReLU(alpha=0.3)(conv1)
conv1 = Dropout(0.2)(conv1)
conv1 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv1)
conv1 = Conv2D(32, (3, 3), dilation_rate=2, padding='same')(conv1)
conv1 = LeakyReLU(alpha=0.3)(conv1)
conv1 = Conv2D(32, (3, 3), dilation_rate=4, padding='same')(conv1)
conv1 = LeakyReLU(alpha=0.3)(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
#pool1 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(pool1)
conv2 = Conv2D(64, (3, 3), padding='same')(
pool1) #,activation='relu', padding='same')(pool1)
conv2 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv2)
conv2 = LeakyReLU(alpha=0.3)(conv2)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3), dilation_rate=2, padding='same')(conv2)
conv2 = LeakyReLU(alpha=0.3)(conv2)
conv2 = Conv2D(64, (3, 3), dilation_rate=4, padding='same')(
conv2) #,W_regularizer=l2(0.01), b_regularizer=l2(0.01))(conv2)
conv2 = LeakyReLU(alpha=0.3)(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
#crop = Cropping2D(cropping=((int(3 * patch_height / 8), int(3 * patch_height / 8)), (int(3 * patch_width / 8), int(3 * patch_width / 8))))(conv1)
#conv3 = concatenate([crop,pool2], axis=1)
conv3 = Conv2D(128, (3, 3), padding='same')(
pool2) #, activation='relu', padding='same')(conv3)
conv3 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv3)
conv3 = LeakyReLU(alpha=0.3)(conv3)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3), dilation_rate=2, padding='same')(
conv3) #,W_regularizer=l2(0.01), b_regularizer=l2(0.01))(conv3)
conv3 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv3)
conv3 = LeakyReLU(alpha=0.3)(conv3)
conv3 = Conv2D(128, (3, 3), dilation_rate=4, padding='same')(conv3)
conv3 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='RandomNormal',
gamma_initializer='one')(conv3)
conv3 = LeakyReLU(alpha=0.3)(conv3)
#up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = concatenate([UpSampling2D(size=(2, 2))(conv3), conv2], axis=1)
conv4 = Conv2D(64, (3, 3), padding='same')(up1)
conv4 = LeakyReLU(alpha=0.3)(conv4)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3), padding='same')(conv4)
conv4 = LeakyReLU(alpha=0.3)(conv4)
#conv4 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(conv4)
#
#up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = concatenate([UpSampling2D(size=(2, 2))(conv4), conv1], axis=1)
conv5 = Conv2D(32, (3, 3), padding='same')(up2)
conv5 = LeakyReLU(alpha=0.3)(conv5)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3), padding='same')(conv5)
conv5 = LeakyReLU(alpha=0.3)(conv5)
conv6 = Conv2D(num_lesion_class + 1, (1, 1), padding='same')(conv5)
conv6 = LeakyReLU(alpha=0.3)(conv6)
#conv6 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(conv6)
#for tensorflow
#conv6 = core.Reshape((patch_height*patch_width,num_lesion_class+1))(conv6)
#for theano
conv6 = core.Reshape(
((num_lesion_class + 1), patch_height * patch_width))(conv6)
conv6 = core.Permute((2, 1))(conv6)
############
act = Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=act)
return model
def test_reshape():
layer = core.Reshape(dims=(10, 10))
_runner(layer)
def R_Unet(n_ch, patch_height, patch_width):
inputs = Input((n_ch, patch_height, patch_width))
conv1 = Conv2D(32, (1, 1), activation=None, padding='same')(inputs)
conv1 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='zero',
gamma_initializer='one')(conv1)
conv1 = Activation('relu')(conv1)
conv1 = DenseBlock(conv1, 32) #48
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
conv2 = DenseBlock(pool1, 64) #24
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
conv3 = DenseBlock(pool2, 64) #12
pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)
conv4 = DenseBlock(pool3, 64) # 12
up1 = Conv2DTranspose(64, (3, 3),
strides=2,
activation='relu',
padding='same')(conv4)
up1 = concatenate([up1, conv3], axis=1)
conv5 = DenseBlock(up1, 64)
up2 = Conv2DTranspose(64, (3, 3),
strides=2,
activation='relu',
padding='same')(conv5)
up2 = concatenate([up2, conv2], axis=1)
conv6 = DenseBlock(up2, 64)
up3 = Conv2DTranspose(64, (3, 3),
strides=2,
activation='relu',
padding='same')(conv6)
up3 = concatenate([up3, conv1], axis=1)
conv7 = DenseBlock(up3, 32)
conv8 = Conv2D(num_lesion_class + 1, (1, 1),
activation='relu',
padding='same')(conv7)
# conv6 = normalization.BatchNormalization(epsilon=1e-06, mode=1, axis=-1, momentum=0.9, weights=None, beta_init='zero', gamma_init='one')(conv6)
# for tensorflow
# conv6 = core.Reshape((patch_height*patch_width,num_lesion_class+1))(conv6)
# for theano
conv8 = core.Reshape(
((num_lesion_class + 1), patch_height * patch_width))(conv8)
conv8 = core.Permute((2, 1))(conv8)
############
act = Activation('softmax')(conv8)
model = Model(inputs=inputs, outputs=act)
return model
def get_unet_seg(n_ch, img_rows=480, img_cols=480):
inputs = Input((n_ch, img_rows, img_cols))
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv1_1")(inputs)
conv1 = BatchNormalization(axis=1, name="conv1_2")(conv1)
conv1 = Dropout(0.5, name="conv1_3")(conv1)
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv1_4")(conv1)
conv1 = BatchNormalization(axis=1, name="conv1_5")(conv1)
conv1.trainable = False
pool1 = MaxPooling2D((2, 2), data_format='channels_first',
name="conv1_6")(conv1)
pool1.trainable = False
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv2_1")(pool1)
conv2 = BatchNormalization(axis=1, name="conv2_2")(conv2)
conv2 = Dropout(0.5, name="conv2_3")(conv2)
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv2_4")(conv2)
conv2 = BatchNormalization(axis=1, name="conv2_5")(conv2)
conv2.trainable = False
pool2 = MaxPooling2D((2, 2), data_format='channels_first',
name="conv2_6")(conv2)
pool2.trainable = False
conv3 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv3_1")(pool2)
conv3 = BatchNormalization(axis=1, name="conv3_2")(conv3)
conv3 = Dropout(0.5, name="conv3_3")(conv3)
conv3 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv3_4")(conv3)
conv3 = BatchNormalization(axis=1, name="conv3_5")(conv3)
conv3.trainable = False
pool3 = MaxPooling2D((2, 2), data_format='channels_first',
name="conv3_6")(conv3)
pool3.trainable = False
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv4_1")(pool3)
conv4 = BatchNormalization(axis=1, name="conv4_2")(conv4)
conv4 = Dropout(0.5, name="conv4_3")(conv4)
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv4_4")(conv4)
conv4 = BatchNormalization(axis=1, name="conv4_5")(conv4)
conv4.trainable = False
pool4 = MaxPooling2D((2, 2), data_format='channels_first',
name="conv4_6")(conv4)
pool4.trainable = False
conv5 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv5_1")(pool4)
conv5 = BatchNormalization(axis=1, name="conv5_2")(conv5)
conv5 = Dropout(0.5, name="conv5_3")(conv5)
conv5 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first',
name="conv5_4")(conv5)
conv5 = BatchNormalization(axis=1, name="conv5_5")(conv5)
conv5.trainable = False
up1 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv5)
up1 = concatenate([conv4, up1], axis=1)
conv6 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up1)
conv6 = BatchNormalization(axis=1)(conv6)
conv6 = Dropout(0.3)(conv6)
conv6 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv6)
conv6 = BatchNormalization(axis=1)(conv6)
up2 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv6)
up2 = concatenate([conv3, up2], axis=1)
conv7 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up2)
conv7 = BatchNormalization(axis=1)(conv7)
conv7 = Dropout(0.3)(conv7)
conv7 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv7)
conv7 = BatchNormalization(axis=1)(conv7)
up3 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv7)
up3 = concatenate([conv2, up3], axis=1)
conv8 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up3)
conv8 = BatchNormalization(axis=1)(conv8)
conv8 = Dropout(0.3)(conv8)
conv8 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv8)
conv8 = BatchNormalization(axis=1)(conv8)
up4 = UpSampling2D(size=(2, 2), data_format='channels_first')(conv8)
up4 = concatenate([conv1, up4], axis=1)
conv9 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(up4)
conv9 = BatchNormalization(axis=1)(conv9)
conv9 = Dropout(0.3)(conv9)
conv9 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_first')(conv9)
conv9 = BatchNormalization(axis=1)(conv9)
conv10 = Conv2D(2, (1, 1),
activation='relu',
padding='same',
data_format='channels_first')(conv9)
conv10 = BatchNormalization(axis=1)(conv10)
conv10 = core.Reshape((2, img_rows * img_cols))(conv10)
conv10 = core.Permute((2, 1))(conv10)
############
conv10 = core.Activation('softmax')(conv10)
model = Model(input=inputs, output=conv10)
adaGrad = Adagrad(lr=1e-7, epsilon=1e-7, decay=1e-6)
model.compile(optimizer='sgd',
loss='categorical_crossentropy',
metrics=['accuracy'])
return model
class LayerCorrectnessTest(keras_parameterized.TestCase):
def setUp(self):
super(LayerCorrectnessTest, self).setUp()
# Set two virtual CPUs to test MirroredStrategy with multiple devices
cpus = tf.config.list_physical_devices('CPU')
tf.config.set_logical_device_configuration(cpus[0], [
tf.config.LogicalDeviceConfiguration(),
tf.config.LogicalDeviceConfiguration(),
])
def _create_model_from_layer(self, layer, input_shapes):
inputs = [layers.Input(batch_input_shape=s) for s in input_shapes]
if len(inputs) == 1:
inputs = inputs[0]
y = layer(inputs)
model = models.Model(inputs, y)
model.compile('sgd', 'mse')
return model
@parameterized.named_parameters(
('LeakyReLU', advanced_activations.LeakyReLU, (2, 2)),
('PReLU', advanced_activations.PReLU, (2, 2)),
('ELU', advanced_activations.ELU, (2, 2)),
('ThresholdedReLU', advanced_activations.ThresholdedReLU, (2, 2)),
('Softmax', advanced_activations.Softmax, (2, 2)),
('ReLU', advanced_activations.ReLU, (2, 2)),
('Conv1D', lambda: convolutional.Conv1D(2, 2), (2, 2, 1)),
('Conv2D', lambda: convolutional.Conv2D(2, 2), (2, 2, 2, 1)),
('Conv3D', lambda: convolutional.Conv3D(2, 2), (2, 2, 2, 2, 1)),
('Conv2DTranspose', lambda: convolutional.Conv2DTranspose(2, 2),
(2, 2, 2, 2)),
('SeparableConv2D', lambda: convolutional.SeparableConv2D(2, 2),
(2, 2, 2, 1)),
('DepthwiseConv2D', lambda: convolutional.DepthwiseConv2D(2, 2),
(2, 2, 2, 1)),
('UpSampling2D', convolutional.UpSampling2D, (2, 2, 2, 1)),
('ZeroPadding2D', convolutional.ZeroPadding2D, (2, 2, 2, 1)),
('Cropping2D', convolutional.Cropping2D, (2, 3, 3, 1)),
('ConvLSTM2D',
lambda: convolutional_recurrent.ConvLSTM2D(4, kernel_size=(2, 2)),
(4, 4, 4, 4, 4)),
('Dense', lambda: core.Dense(2), (2, 2)),
('Dropout', lambda: core.Dropout(0.5), (2, 2)),
('SpatialDropout2D', lambda: core.SpatialDropout2D(0.5), (2, 2, 2, 2)),
('Activation', lambda: core.Activation('sigmoid'), (2, 2)),
('Reshape', lambda: core.Reshape((1, 4, 1)), (2, 2, 2)),
('Permute', lambda: core.Permute((2, 1)), (2, 2, 2)),
('Attention', dense_attention.Attention, [(2, 2, 3), (2, 3, 3),
(2, 3, 3)]),
('AdditiveAttention', dense_attention.AdditiveAttention, [(2, 2, 3),
(2, 3, 3),
(2, 3, 3)]),
('Embedding', lambda: embeddings.Embedding(4, 4),
(2, 4), 2e-3, 2e-3, np.random.randint(4, size=(2, 4))),
('LocallyConnected1D', lambda: local.LocallyConnected1D(2, 2),
(2, 2, 1)),
('LocallyConnected2D', lambda: local.LocallyConnected2D(2, 2),
(2, 2, 2, 1)),
('Add', merge.Add, [(2, 2), (2, 2)]),
('Subtract', merge.Subtract, [(2, 2), (2, 2)]),
('Multiply', merge.Multiply, [(2, 2), (2, 2)]),
('Average', merge.Average, [(2, 2), (2, 2)]),
('Maximum', merge.Maximum, [(2, 2), (2, 2)]),
('Minimum', merge.Minimum, [(2, 2), (2, 2)]),
('Concatenate', merge.Concatenate, [(2, 2), (2, 2)]),
('Dot', lambda: merge.Dot(1), [(2, 2), (2, 2)]),
('GaussianNoise', lambda: noise.GaussianNoise(0.5), (2, 2)),
('GaussianDropout', lambda: noise.GaussianDropout(0.5), (2, 2)),
('AlphaDropout', lambda: noise.AlphaDropout(0.5), (2, 2)),
('BatchNormalization', normalization_v2.BatchNormalization,
(2, 2), 1e-2, 1e-2),
('LayerNormalization', normalization.LayerNormalization, (2, 2)),
('LayerNormalizationUnfused',
lambda: normalization.LayerNormalization(axis=1), (2, 2, 2)),
('MaxPooling2D', pooling.MaxPooling2D, (2, 2, 2, 1)),
('AveragePooling2D', pooling.AveragePooling2D, (2, 2, 2, 1)),
('GlobalMaxPooling2D', pooling.GlobalMaxPooling2D, (2, 2, 2, 1)),
('GlobalAveragePooling2D', pooling.GlobalAveragePooling2D,
(2, 2, 2, 1)),
('SimpleRNN', lambda: recurrent.SimpleRNN(units=4),
(4, 4, 4), 1e-2, 1e-2),
('GRU', lambda: recurrent.GRU(units=4), (4, 4, 4)),
('LSTM', lambda: recurrent.LSTM(units=4), (4, 4, 4)),
('GRUV2', lambda: recurrent_v2.GRU(units=4), (4, 4, 4)),
('LSTMV2', lambda: recurrent_v2.LSTM(units=4), (4, 4, 4)),
('TimeDistributed', lambda: wrappers.TimeDistributed(core.Dense(2)),
(2, 2, 2)),
('Bidirectional',
lambda: wrappers.Bidirectional(recurrent.SimpleRNN(units=4)),
(2, 2, 2)),
('AttentionLayerCausal',
lambda: dense_attention.Attention(causal=True), [(2, 2, 3), (2, 3, 3),
(2, 3, 3)]),
('AdditiveAttentionLayerCausal',
lambda: dense_attention.AdditiveAttention(causal=True), [(2, 3, 4),
(2, 3, 4),
(2, 3, 4)]),
)
def test_layer(self,
f32_layer_fn,
input_shape,
rtol=2e-3,
atol=2e-3,
input_data=None):
"""Tests a layer by comparing the float32 and mixed precision weights.
A float32 layer, a mixed precision layer, and a distributed mixed precision
layer are run. The three layers are identical other than their dtypes and
distribution strategies. The outputs after predict() and weights after fit()
are asserted to be close.
Args:
f32_layer_fn: A function returning a float32 layer. The other two layers
will automatically be created from this
input_shape: The shape of the input to the layer, including the batch
dimension. Or a list of shapes if the layer takes multiple inputs.
rtol: The relative tolerance to be asserted.
atol: The absolute tolerance to be asserted.
input_data: A Numpy array with the data of the input. If None, input data
will be randomly generated
"""
if f32_layer_fn == convolutional.ZeroPadding2D and \
tf.test.is_built_with_rocm():
return
if isinstance(input_shape[0], int):
input_shapes = [input_shape]
else:
input_shapes = input_shape
strategy = create_mirrored_strategy()
f32_layer = f32_layer_fn()
# Create the layers
assert f32_layer.dtype == f32_layer._compute_dtype == 'float32'
config = f32_layer.get_config()
config['dtype'] = policy.Policy('mixed_float16')
mp_layer = f32_layer.__class__.from_config(config)
distributed_mp_layer = f32_layer.__class__.from_config(config)
# Compute per_replica_input_shapes for the distributed model
global_batch_size = input_shapes[0][0]
assert global_batch_size % strategy.num_replicas_in_sync == 0, (
'The number of replicas, %d, does not divide the global batch size of '
'%d' % (strategy.num_replicas_in_sync, global_batch_size))
per_replica_batch_size = (global_batch_size //
strategy.num_replicas_in_sync)
per_replica_input_shapes = [(per_replica_batch_size, ) + s[1:]
for s in input_shapes]
# Create the models
f32_model = self._create_model_from_layer(f32_layer, input_shapes)
mp_model = self._create_model_from_layer(mp_layer, input_shapes)
with strategy.scope():
distributed_mp_model = self._create_model_from_layer(
distributed_mp_layer, per_replica_input_shapes)
# Set all model weights to the same values
f32_weights = f32_model.get_weights()
mp_model.set_weights(f32_weights)
distributed_mp_model.set_weights(f32_weights)
# Generate input data
if input_data is None:
# Cast inputs to float16 to avoid measuring error from having f16 layers
# cast to float16.
input_data = [
np.random.normal(size=s).astype('float16')
for s in input_shapes
]
if len(input_data) == 1:
input_data = input_data[0]
# Assert all models have close outputs.
f32_output = f32_model.predict(input_data)
mp_output = mp_model.predict(input_data)
self.assertAllClose(mp_output, f32_output, rtol=rtol, atol=atol)
self.assertAllClose(distributed_mp_model.predict(input_data),
f32_output,
rtol=rtol,
atol=atol)
# Run fit() on models
output = np.random.normal(
size=f32_model.outputs[0].shape).astype('float16')
for model in f32_model, mp_model, distributed_mp_model:
model.fit(input_data, output, batch_size=global_batch_size)
# Assert all models have close weights
f32_weights = f32_model.get_weights()
self.assertAllClose(mp_model.get_weights(),
f32_weights,
rtol=rtol,
atol=atol)
self.assertAllClose(distributed_mp_model.get_weights(),
f32_weights,
rtol=rtol,
atol=atol)
def baseline_unet(patch_height, patch_width, n_ch):
inputs = Input(shape=(patch_height, patch_width, n_ch))
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(conv1)
pool1 = MaxPooling2D((2, 2), data_format='channels_last')(conv1)
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(conv2)
pool2 = MaxPooling2D((2, 2), data_format='channels_last')(conv2)
conv3 = Conv2D(128, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(conv3)
up1 = UpSampling2D(size=(2, 2), data_format='channels_last')(conv3)
up1 = Concatenate(axis=3)([conv2, up1])
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(conv4)
up2 = UpSampling2D(size=(2, 2), data_format='channels_last')(conv4)
up2 = Concatenate(axis=3)([conv1, up2])
conv5 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3),
activation='relu',
padding='same',
data_format='channels_last')(conv5)
conv6 = Conv2D(2, (1, 1),
activation='relu',
padding='same',
data_format='channels_last')(conv5)
conv6 = core.Reshape((patch_height * patch_width, 2))(conv6)
conv7 = core.Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=conv7)
print(model.summary())
opt = Adam(lr=0.01, beta_1=0.9, beta_2=0.999, epsilon=0.1, decay=0.0)
model.compile(optimizer=opt,
loss='categorical_crossentropy',
metrics=['accuracy'])
return model
def MultiCNN(trainX,
trainY,
valX=None,
valY=None,
batch_size=1200,
nb_epoch=500,
earlystop=None,
transferlayer=1,
weights=None,
forkinas=False,
compiletimes=0,
compilemodels=None,
predict=False):
input_row = trainX.shape[2]
input_col = trainX.shape[3]
trainX_t = trainX
valX_t = valX
if (earlystop is not None):
early_stopping = EarlyStopping(monitor='val_loss', patience=earlystop)
nb_epoch = 10000
#set to a very big value since earlystop used
trainX_t.shape = (trainX_t.shape[0], input_row, input_col)
if (valX is not None):
valX_t.shape = (valX_t.shape[0], input_row, input_col)
if compiletimes == 0:
input = Input(shape=(input_row, input_col))
filtersize1 = 1
filtersize2 = 9
filtersize3 = 10
filter1 = 200
filter2 = 150
filter3 = 200
dropout1 = 0.75
dropout2 = 0.75
dropout4 = 0.75
dropout5 = 0.75
dropout6 = 0
L1CNN = 0
nb_classes = 2
batch_size = 1200
actfun = "relu"
optimization = 'adam'
attentionhidden_x = 10
attentionhidden_xr = 8
attention_reg_x = 0.151948
attention_reg_xr = 2
dense_size1 = 149
dense_size2 = 8
dropout_dense1 = 0.298224
dropout_dense2 = 0
input = Input(shape=(input_row, input_col))
x = conv.Convolution1D(filter1,
filtersize1,
init='he_normal',
W_regularizer=l1(L1CNN),
border_mode="same")(input)
x = Dropout(dropout1)(x)
x = Activation(actfun)(x)
x = conv.Convolution1D(filter2,
filtersize2,
init='he_normal',
W_regularizer=l1(L1CNN),
border_mode="same")(x)
x = Dropout(dropout2)(x)
x = Activation(actfun)(x)
x = conv.Convolution1D(filter3,
filtersize3,
init='he_normal',
W_regularizer=l1(L1CNN),
border_mode="same")(x)
x = Activation(actfun)(x)
x_reshape = core.Reshape((x._keras_shape[2], x._keras_shape[1]))(x)
x = Dropout(dropout4)(x)
x_reshape = Dropout(dropout5)(x_reshape)
decoder_x = Attention(hidden=attentionhidden_x,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_x)) # success
decoded_x = decoder_x(x)
output_x = myFlatten(x._keras_shape[2])(decoded_x)
decoder_xr = Attention(hidden=attentionhidden_xr,
activation='linear',
init='he_normal',
W_regularizer=l1(attention_reg_xr))
decoded_xr = decoder_xr(x_reshape)
output_xr = myFlatten(x_reshape._keras_shape[2])(decoded_xr)
output = merge([output_x, output_xr], mode='concat')
output = Dropout(dropout6)(output)
output = Dense(dense_size1, init='he_normal',
activation='relu')(output)
output = Dropout(dropout_dense1)(output)
output = Dense(dense_size2, activation="relu",
init='he_normal')(output)
output = Dropout(dropout_dense2)(output)
out = Dense(nb_classes, init='he_normal', activation='softmax')(output)
cnn = Model(input, out)
cnn.compile(loss='binary_crossentropy',
optimizer=optimization,
metrics=['accuracy'])
else:
cnn = compilemodels
if (predict is False):
if (weights is not None and compiletimes == 0): #for the first time
print "load weights:" + weights
if not forkinas:
cnn.load_weights(weights)
else:
cnn2 = copy.deepcopy(cnn)
cnn2.load_weights(weights)
for l in range((len(cnn2.layers) -
transferlayer)): #the last cnn is not included
cnn.layers[l].set_weights(cnn2.layers[l].get_weights())
#cnn.layers[l].trainable= False # for frozen layer
if (valX is not None):
if (earlystop is None):
fitHistory = cnn.fit(trainX_t,
trainY,
batch_size=batch_size,
nb_epoch=nb_epoch,
validation_data=(valX_t, valY))
else:
fitHistory = cnn.fit(trainX_t,
trainY,
batch_size=batch_size,
nb_epoch=nb_epoch,
validation_data=(valX_t, valY),
callbacks=[early_stopping])
else:
fitHistory = cnn.fit(trainX_t,
trainY,
batch_size=batch_size,
nb_epoch=nb_epoch)
return cnn
def unet_model_MultiScale():
inputs = Input(config["input_shape"])
conv1 = Conv3D(32, (3, 3, 3), activation='relu', padding='same')(inputs)
conv1 = Conv3D(32, (3, 3, 3), padding='same')(conv1)
conv1 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='zero',
gamma_initializer='one')(conv1)
conv1 = core.Activation('relu')(conv1)
pool1 = MaxPooling3D(pool_size=config["pool_size"])(conv1)
conv2 = Conv3D(64, (3, 3, 3), activation='relu', padding='same')(pool1)
conv2 = Conv3D(64, (3, 3, 3), padding='same')(conv2)
conv2 = normalization.BatchNormalization(epsilon=2e-05,
axis=1,
momentum=0.9,
weights=None,
beta_initializer='zero',
gamma_initializer='one')(conv2)
conv2 = core.Activation('relu')(conv2)
pool2_1 = MaxPooling3D(pool_size=config["pool_size"])(conv2)
conv3_1 = Conv3D(128, (3, 3, 3), activation='relu',
padding='same')(pool2_1)
conv3_1 = Conv3D(128, (3, 3, 3), activation='relu',
padding='same')(conv3_1)
pool2_2 = MaxPooling3D(pool_size=(4, 4, 4))(conv2)
conv3_2 = Conv3D(128, (3, 3, 3), activation='relu',
padding='same')(pool2_2)
conv3_2 = Conv3D(128, (3, 3, 3), activation='relu',
padding='same')(conv3_2)
fuse = concatenate(
[UpSampling3D(size=config["pool_size"])(conv3_2), conv3_1], axis=1)
conv3_f = Conv3D(128, (3, 3, 3), activation='relu', padding='same')(fuse)
up4 = concatenate([UpSampling3D(size=config["pool_size"])(conv3_f), conv2],
axis=1)
conv4 = Conv3D(64, (3, 3, 3), activation='relu', padding='same')(up4)
conv4 = Conv3D(64, (3, 3, 3), activation='relu', padding='same')(conv4)
up5 = concatenate([UpSampling3D(size=config["pool_size"])(conv4), conv1],
axis=1)
conv5 = Conv3D(32, (3, 3, 3), activation='relu', padding='valid')(up5)
conv5 = Conv3D(32, (3, 3, 3), activation='relu', padding='valid')(conv5)
conv6 = Conv3D(config["n_labels"], (1, 1, 1))(conv5)
conv6 = core.Reshape((1, out_w * out_w * out_w))(conv6)
conv6 = core.Permute((2, 1))(conv6)
#conv6 =
act = Activation('sigmoid')(conv6)
model = Model(inputs=inputs, outputs=act)
#model.compile(optimizer=Adam(lr=config["initial_learning_rate"]), loss='categorical_crossentropy',metrics=['fbeta_score'])
model.compile(optimizer=Adam(lr=config["initial_learning_rate"]),
loss=dice_coef_loss,
metrics=[dice_coef])
return model
def build_model(self):
inputs = Input((self.patch_height, self.patch_width, 3))
block1 = self._ResBlock(inputs, 32) #128 128 32
pool1 = MaxPooling2D(strides=(2, 2))(block1)
block2 = self._ResBlock(pool1, 32) # 64 64 32
pool2 = MaxPooling2D(strides=(2, 2))(block2)
block3 = self._ResBlock(pool2, 64) # 32 32 64
pool3 = MaxPooling2D(strides=(2, 2))(block3)
block4_1 = self._ResBlock(pool3, 128, 1) #16 16 128
block4_2 = self._ResBlock(block4_1, 128, 2)
block4 = self._ResBlock(block4_2, 128, 4)
aspp = self._aspp(block4)
decode_16 = Conv2DTranspose(64, (3, 3),
strides=(4, 4),
activation="relu",
padding="same")(block4)
decode_32 = Conv2DTranspose(64, (3, 3),
strides=(2, 2),
activation="relu",
padding="same")(block3)
up1 = concatenate([block2, decode_16, decode_32], axis=3)
aspp = Conv2DTranspose(32, (3, 3),
strides=(4, 4),
activation="relu",
padding="same")(aspp)
up2 = concatenate([aspp, up1], axis=3)
decode = Conv2D(32, (3, 3), strides=(1, 1), padding="same")(up2)
decode = BatchNormalization()(decode)
decode = Activation('relu')(decode)
decode2 = Conv2DTranspose(32, (3, 3),
strides=(2, 2),
activation="relu",
padding="same")(decode)
decode3 = concatenate([decode2, block1], axis=3)
conv8 = Conv2D(self.num_seg_class + 1, (1, 1),
activation='relu',
padding='same')(decode3)
conv8 = core.Reshape((self.patch_height * self.patch_width,
(self.num_seg_class + 1)))(conv8)
############
act = Activation('softmax')(conv8)
model = Model(inputs=inputs, outputs=act)
model.compile(optimizer='adam',
loss='categorical_crossentropy',
metrics=['categorical_accuracy'])
plot_model(model,
to_file=os.path.join(self.config.checkpoint,
"deeplabv3+.png"),
show_shapes=True)
self.model = model
def get_gnet(n_ch,patch_height,patch_width):
inputs = Input((n_ch, patch_height, patch_width))
conv1 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv1)
up1 = UpSampling2D(size=(2, 2))(conv1)
#
conv2 = Convolution2D(16, 3, 3, activation='relu', border_mode='same')(up1)
conv2 = Dropout(0.2)(conv2)
conv2 = Convolution2D(16, 3, 3, activation='relu', border_mode='same')(conv2)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv2)
#
conv3 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(pool1)
conv3 = Dropout(0.2)(conv3)
conv3 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv3)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv3)
#
conv4 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(pool2)
conv4 = Dropout(0.2)(conv4)
conv4 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(conv4)
pool3 = MaxPooling2D(pool_size=(2, 2))(conv4)
#
conv5 = Convolution2D(128, 3, 3, activation='relu', border_mode='same')(pool3)
conv5 = Dropout(0.2)(conv5)
conv5 = Convolution2D(128, 3, 3, activation='relu', border_mode='same')(conv5)
#
up2 = concatenate([UpSampling2D(size=(2, 2))(conv5), conv4], axis=3)
conv6 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(up2)
conv6 = Dropout(0.2)(conv6)
conv6 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(conv6)
#
up3 = concatenate([UpSampling2D(size=(2, 2))(conv6), conv3], axis=3)
conv7 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(up3)
conv7 = Dropout(0.2)(conv7)
conv7 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv7)
#
up4 = concatenate([UpSampling2D(size=(2, 2))(conv7), conv2], axis=3)
conv8 = Convolution2D(16, 3, 3, activation='relu', border_mode='same')(up4)
conv8 = Dropout(0.2)(conv8)
conv8 = Convolution2D(16, 3, 3, activation='relu', border_mode='same')(conv8)
#
pool4 = MaxPooling2D(pool_size=(2, 2))(conv8)
conv9 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(pool4)
conv9 = Dropout(0.2)(conv9)
conv9 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(conv9)
#
conv10 = Convolution2D(2, 1, 1, activation='relu', border_mode='same')(conv9)
conv10 = core.Reshape((2,patch_height*patch_width))(conv10)
conv10 = core.Permute((2,1))(conv10)
############
conv10 = core.Activation('softmax')(conv10)
model = Model(input=inputs, output=conv10)
# sgd = SGD(lr=0.01, decay=1e-6, momentum=0.3, nesterov=False)
model.compile(optimizer='sgd', loss='categorical_crossentropy',metrics=['accuracy'])
x, y = imageSegmentationGenerator(cfg.train_images, cfg.train_annotations, cfg.train_batch_size,
cfg.n_classes, cfg.input_height, cfg.input_width, cfg.output_height,
cfg.output_width)
model.fit(
x, y,
steps_per_epoch=int(cfg.train_data_number / cfg.train_batch_size),
max_queue_size=8, workers=4, validation_data=5, epochs=cfg.epochs
)
# return model
def get_dilated_unet(n_ch, patch_height, patch_width, dilaterate=3):
inputs = Input(shape=(n_ch, patch_height, patch_width))
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same', dilation_rate=dilaterate,
data_format='channels_first')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Conv2D(32, (3, 3), activation='relu', padding='same', dilation_rate=dilaterate,
data_format='channels_first')(conv1)
pool1 = MaxPooling2D((2, 2))(conv1)
#
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same', dilation_rate=dilaterate,
data_format='channels_first')(pool1)
conv2 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3, 3), activation='relu', padding='same', dilation_rate=dilaterate,
data_format='channels_first')(conv2)
pool2 = MaxPooling2D((2, 2))(conv2)
#
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same', dilation_rate=dilaterate,
data_format='channels_first')(pool2)
conv3 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3, 3), activation='relu', padding='same', dilation_rate=dilaterate,
data_format='channels_first')(conv3)
up1 = UpSampling2D(size=(2, 2))(conv3)
up1 = concatenate([conv2, up1], axis=1)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same', data_format='channels_first')(up1)
conv4 = Dropout(0.2)(conv4)
conv4 = Conv2D(64, (3, 3), activation='relu', padding='same', data_format='channels_first')(conv4)
#
up2 = UpSampling2D(size=(2, 2))(conv4)
up2 = concatenate([conv1, up2], axis=1)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same', data_format='channels_first')(up2)
conv5 = Dropout(0.2)(conv5)
conv5 = Conv2D(32, (3, 3), activation='relu', padding='same', data_format='channels_first')(conv5)
#
conv6 = Conv2D(2, (1, 1), activation='relu', padding='same', data_format='channels_first')(conv5)
conv6 = core.Reshape((2, patch_height * patch_width))(conv6)
conv6 = core.Permute((2, 1))(conv6)
############
conv7 = core.Activation('softmax')(conv6)
model = Model(inputs=inputs, outputs=conv7)
# scheduler = LearningRateScheduler(mlr.lr_scheduler)
sgd = SGD(lr=0.01, decay=2e-5, momentum=0.8, nesterov=False)
model.compile(optimizer=sgd, loss='binary_crossentropy', metrics=['accuracy'])
# adam=optimizers.Adam(lr=0.01, beta_1=0.9, beta_2=0.999, epsilon=1e-07)
# model.compile(optimizer=adam, loss='categorical_crossentropy', metrics=['accuracy'])
# 1、目标函数
# (1）mean_squared_error / mse 均方误差，常用的目标函数，公式为((y_pred-y_true) ** 2).mean()
# (2）mean_absolute_error / mae绝对值均差，公式为( | y_pred - y_true |).mean()
# (3)mean_absolute_percentage_error / mape公式为：(| (y_true - y_pred) / clip((| y_true |), epsilon, infinite) |).mean(axis=-1) * 100，和mae的区别就是，累加的是（预测值与实际值的差）除以（剔除不介于epsilon和infinite之间的实际值)，然后求均值。
# (4)mean_squared_logarithmic_error / msle公式为： (log(clip(y_pred, epsilon, infinite) + 1) - log(clip(y_true, epsilon, infinite) + 1.)) ^ 2.mean(axis=-1)，这个就是加入了log对数，剔除不介于epsilon和infinite之间的预测值与实际值之后，然后取对数，作差，平方，累加求均值。
# (5)squared_hinge公式为：(max(1 - y_truey_pred, 0)) ^ 2.mean(axis=-1)，取1减去预测值与实际值乘积的结果与0比相对大的值的平方的累加均值。
# (6)hinge公式为：(max(1 - y_truey_pred, 0)).mean(axis=-1)，取1减去预测值与实际值乘积的结果与0比相对大的值的的累加均值。
# (7)binary_crossentropy: 常说的逻辑回归, 就是常用的交叉熵函
# (8)categorical_crossentropy: 多分类的逻辑
#
# 2、性能评估函数：
# (1)binary_accuracy: 对二分类问题, 计算在所有预测值上的平均正确率
# (2)categorical_accuracy: 对多分类问题, 计算再所有预测值上的平均正确率
# (3)sparse_categorical_accuracy: 与categorical_accuracy相同, 在对稀疏的目标值预测时有用
# (4)top_k_categorical_accracy: 计算top - k正确率, 当预测值的前k个值中存在目标类别即认为预测正确
# (5)sparse_top_k_categorical_accuracy：与top_k_categorical_accracy作用相同，但适用于稀疏情况
return model
def get_gnet(n_ch, patch_height, patch_width):
inputs = Input((n_ch, patch_height, patch_width))
conv1 = Convolution2D(32, 3, 3, activation='relu',
border_mode='same')(inputs)
conv1 = Dropout(0.2)(conv1)
conv1 = Convolution2D(32, 3, 3, activation='relu',
border_mode='same')(conv1)
up1 = UpSampling2D(size=(2, 2))(conv1)
#
conv2 = Convolution2D(16, 3, 3, activation='relu', border_mode='same')(up1)
conv2 = Dropout(0.2)(conv2)
conv2 = Convolution2D(16, 3, 3, activation='relu',
border_mode='same')(conv2)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv2)
#
conv3 = Convolution2D(32, 3, 3, activation='relu',
border_mode='same')(pool1)
conv3 = Dropout(0.2)(conv3)
conv3 = Convolution2D(32, 3, 3, activation='relu',
border_mode='same')(conv3)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv3)
#
conv4 = Convolution2D(64, 3, 3, activation='relu',
border_mode='same')(pool2)
conv4 = Dropout(0.2)(conv4)
conv4 = Convolution2D(64, 3, 3, activation='relu',
border_mode='same')(conv4)
pool3 = MaxPooling2D(pool_size=(2, 2))(conv4)
#
conv5 = Convolution2D(128, 3, 3, activation='relu',
border_mode='same')(pool3)
conv5 = Dropout(0.2)(conv5)
conv5 = Convolution2D(128, 3, 3, activation='relu',
border_mode='same')(conv5)
#
up2 = merge([UpSampling2D(size=(2, 2))(conv5), conv4],
mode='concat',
concat_axis=1)
conv6 = Convolution2D(64, 3, 3, activation='relu', border_mode='same')(up2)
conv6 = Dropout(0.2)(conv6)
conv6 = Convolution2D(64, 3, 3, activation='relu',
border_mode='same')(conv6)
#
up3 = merge([UpSampling2D(size=(2, 2))(conv6), conv3],
mode='concat',
concat_axis=1)
conv7 = Convolution2D(32, 3, 3, activation='relu', border_mode='same')(up3)
conv7 = Dropout(0.2)(conv7)
conv7 = Convolution2D(32, 3, 3, activation='relu',
border_mode='same')(conv7)
#
up4 = merge([UpSampling2D(size=(2, 2))(conv7), conv2],
mode='concat',
concat_axis=1)
conv8 = Convolution2D(16, 3, 3, activation='relu', border_mode='same')(up4)
conv8 = Dropout(0.2)(conv8)
conv8 = Convolution2D(16, 3, 3, activation='relu',
border_mode='same')(conv8)
#
pool4 = MaxPooling2D(pool_size=(2, 2))(conv8)
conv9 = Convolution2D(32, 3, 3, activation='relu',
border_mode='same')(pool4)
conv9 = Dropout(0.2)(conv9)
conv9 = Convolution2D(32, 3, 3, activation='relu',
border_mode='same')(conv9)
#
conv10 = Convolution2D(2, 1, 1, activation='relu',
border_mode='same')(conv9)
conv10 = core.Reshape((2, patch_height * patch_width))(conv10)
conv10 = core.Permute((2, 1))(conv10)
############
conv10 = core.Activation('softmax')(conv10)
model = Model(input=inputs, output=conv10)
# sgd = SGD(lr=0.01, decay=1e-6, momentum=0.3, nesterov=False)
model.compile(optimizer='sgd',
loss='categorical_crossentropy',
metrics=['accuracy'])
return model
def build_model(self):
inputs = Input((self.patch_height, self.patch_width, 3))
conv = Conv2D(16, (3, 3), padding='same')(inputs) # 'valid'
conv = LeakyReLU(alpha=0.3)(conv)
conv1 = self.UnetConv2D(conv, 32, is_batchnorm=True) # 32 128
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
conv2 = self.UnetConv2D(pool1, 32, is_batchnorm=True) # 32 64
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
conv3 = self.UnetConv2D(pool2, 64, is_batchnorm=True) # 64 32
pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)
conv4 = self.UnetConv2D(pool3, 64, is_batchnorm=True) # 64 16
pool4 = MaxPooling2D(pool_size=(2, 2))(conv4)
center = self.UnetConv2D(pool4, 128, is_batchnorm=True) # 128 8
gating = self.UnetGatingSignal(center, is_batchnorm=True)
attn_1 = self.AttnGatingBlock(conv4, gating, 128)
up1 = concatenate([
Conv2DTranspose(64, (3, 3), strides=(2, 2),
padding='same')(center), attn_1
],
axis=3)
attn_2 = self.AttnGatingBlock(conv3, gating, 64)
up2 = concatenate([
Conv2DTranspose(64, (3, 3), strides=(2, 2), padding='same')(up1),
attn_2
],
axis=3)
attn_3 = self.AttnGatingBlock(conv2, gating, 32)
up3 = concatenate([
Conv2DTranspose(32, (3, 3), strides=(2, 2), padding='same')(up2),
attn_3
],
axis=3)
up4 = concatenate([
Conv2DTranspose(32,
(3, 3), strides=(2, 2), padding='same')(up3), conv1
],
axis=3)
conv8 = Conv2D(self.num_seg_class + 1, (1, 1),
activation='relu',
padding='same')(up4)
conv8 = core.Reshape((self.patch_height * self.patch_width,
(self.num_seg_class + 1)))(conv8)
############
act = Activation('softmax')(conv8)
model = Model(inputs=inputs, outputs=act)
model.compile(optimizer='adam',
loss='categorical_crossentropy',
metrics=['categorical_accuracy'])
plot_model(model,
to_file=os.path.join(self.config.checkpoint, "model.png"),
show_shapes=True)
self.model = model