def sam_resnet(x):
# Dilated Convolutional Network
dcn = dcn_resnet(input_tensor=x[0])
conv_feat = Convolution2D(512, 3, 3, border_mode='same', activation='relu')(dcn.output)
# Attentive Convolutional LSTM
att_convlstm = Lambda(repeat, repeat_shape)(conv_feat)
att_convlstm = AttentiveConvLSTM(nb_filters_in=512, nb_filters_out=512, nb_filters_att=512,
nb_cols=3, nb_rows=3)(att_convlstm)
# Learned Prior (1)
priors1 = LearningPrior(nb_gaussian=nb_gaussian, init=gaussian_priors_init)(x[1])
concateneted = merge([att_convlstm, priors1], mode='concat', concat_axis=1)
learned_priors1 = AtrousConvolution2D(512, 5, 5, border_mode='same', activation='relu',
atrous_rate=(4, 4))(concateneted)
# Learned Prior (2)
priors2 = LearningPrior(nb_gaussian=nb_gaussian, init=gaussian_priors_init)(x[1])
concateneted = merge([learned_priors1, priors2], mode='concat', concat_axis=1)
learned_priors2 = AtrousConvolution2D(512, 5, 5, border_mode='same', activation='relu',
atrous_rate=(4, 4))(concateneted)
# Final Convolutional Layer
outs = Convolution2D(1, 1, 1, border_mode='same', activation='relu')(learned_priors2)
outs_up = Lambda(upsampling, upsampling_shape)(outs)
return [outs_up, outs_up, outs_up]
def sam_resnet(data):
# dcn = dcn_resnet(input_tensor=data, trainable=True)
bn_axis = 3
trainable = True #
# conv_1
conv_1_out = ZeroPadding2D((3, 3), batch_size=1)(data)
conv_1_out = Conv2D(64, (7, 7),
strides=(2, 2),
name='conv1',
trainable=trainable)(conv_1_out)
conv_1_out = BatchNormalization(axis=bn_axis,
name='bn_conv1',
trainable=trainable)(conv_1_out)
conv_1_out_b = Activation('relu')(conv_1_out)
conv_1_out = MaxPooling2D((3, 3), strides=(2, 2),
padding='same')(conv_1_out_b)
# conv_2
conv_2_out = conv_block(conv_1_out,
3, [64, 64, 256],
stage=2,
block='a',
strides=(1, 1),
trainable=trainable)
conv_2_out = identity_block(conv_2_out,
3, [64, 64, 256],
stage=2,
block='b',
trainable=trainable)
conv_2_out = identity_block(conv_2_out,
3, [64, 64, 256],
stage=2,
block='c',
trainable=trainable)
# conv_3
conv_3_out = conv_block(conv_2_out,
3, [128, 128, 512],
stage=3,
block='a',
strides=(2, 2),
trainable=trainable)
conv_3_out = identity_block(conv_3_out,
3, [128, 128, 512],
stage=3,
block='b',
trainable=trainable)
conv_3_out = identity_block(conv_3_out,
3, [128, 128, 512],
stage=3,
block='c',
trainable=trainable)
conv_3_out = identity_block(conv_3_out,
3, [128, 128, 512],
stage=3,
block='d',
trainable=trainable)
# conv_4
conv_4_out = conv_block(conv_3_out,
3, [256, 256, 1024],
stage=4,
block='a',
trainable=trainable)
conv_4_out = identity_block(conv_4_out,
3, [256, 256, 1024],
stage=4,
block='b',
trainable=trainable)
conv_4_out = identity_block(conv_4_out,
3, [256, 256, 1024],
stage=4,
block='c',
trainable=trainable)
conv_4_out = identity_block(conv_4_out,
3, [256, 256, 1024],
stage=4,
block='d',
trainable=trainable)
conv_4_out = identity_block(conv_4_out,
3, [256, 256, 1024],
stage=4,
block='e',
trainable=trainable)
conv_4_out = identity_block(conv_4_out,
3, [256, 256, 1024],
stage=4,
block='f',
trainable=trainable)
# conv_5
conv_5_out = conv_block(conv_4_out,
3, [512, 512, 2048],
stage=5,
block='a',
strides=(1, 1),
trainable=trainable) #
conv_5_out = identity_block(conv_5_out,
3, [512, 512, 2048],
stage=5,
block='b',
trainable=trainable)
conv_5_out = identity_block(conv_5_out,
3, [512, 512, 2048],
stage=5,
block='c',
trainable=trainable)
#
# processing Resnet output
resnet_outs = Conv2D(512, (3, 3),
padding='same',
activation='relu',
name='resnet_out',
trainable=trainable)(conv_5_out)
resnet_outs = Flatten()(resnet_outs)
resnet_outs = RepeatVector(nb_timestep)(resnet_outs)
resnet_outs = Reshape((nb_timestep, 14, 14, 512))(resnet_outs)
# Attentive Convolutional LSTM
convLSTM_outs = AttentiveConvLSTM(nb_filters_in=512,
nb_filters_out=512,
nb_filters_att=512,
nb_cols=3,
nb_rows=3,
name='attenconvsltm')(
resnet_outs) #, trainable=True
convLSTM_outs = Lambda(Kreshape,
arguments={'shape': [-1, 14, 14, 512]},
output_shape=[14, 14, 512])(convLSTM_outs)
# final land output
land_outs = Conv2D(8, (1, 1),
padding='same',
activation='sigmoid',
name='land_con5',
trainable=True)(convLSTM_outs)
# upsamping land output for 3rd block
conv_3_out = Conv2D(64, (3, 3),
padding='same',
activation='relu',
name='conv_3_out',
trainable=True)(conv_3_out)
up3_land_outs = UpSampling2D(size=(2, 2))(land_outs)
up3_land_outs = Concatenate()([conv_3_out, up3_land_outs])
up3_land_outs = Flatten()(up3_land_outs)
up3_land_outs = RepeatVector(nb_timestep)(up3_land_outs)
up3_land_outs = Reshape((nb_timestep, 28, 28, 72))(up3_land_outs)
up3_land_outs = (ConvLSTM2D(filters=8,
kernel_size=(3, 3),
padding='same',
activation='sigmoid',
return_sequences=False,
stateful=False,
name='land_con3'))(up3_land_outs)
# # upsamping land output for 2nd block
conv_2_out = Conv2D(64, (3, 3),
padding='same',
activation='relu',
name='conv_2_out',
trainable=True)(conv_2_out)
up2_land_outs = UpSampling2D(size=(2, 2))(up3_land_outs)
up2_land_outs = Concatenate()([conv_2_out, up2_land_outs])
up2_land_outs = Flatten()(up2_land_outs)
up2_land_outs = RepeatVector(nb_timestep)(up2_land_outs)
up2_land_outs = Reshape((nb_timestep, 56, 56, 72))(up2_land_outs)
up2_land_outs = (ConvLSTM2D(filters=8,
kernel_size=(3, 3),
padding='same',
activation='sigmoid',
return_sequences=False,
stateful=False,
name='land_con2'))(up2_land_outs)
# # # upsamping land output for 1st block
up1_land_outs = UpSampling2D(size=(2, 2))(up2_land_outs)
up1_land_outs = Concatenate()([conv_1_out_b, up1_land_outs])
up1_land_outs = Flatten()(up1_land_outs)
up1_land_outs = RepeatVector(nb_timestep)(up1_land_outs)
up1_land_outs = Reshape((nb_timestep, 112, 112, 72))(up1_land_outs)
up1_land_outs = (ConvLSTM2D(filters=8,
kernel_size=(3, 3),
padding='same',
activation='sigmoid',
return_sequences=False,
stateful=False,
name='land_con1'))(up1_land_outs)
# outs = Lambda(Kpool, arguments={'pool_size': (14, 14)}, output_shape=[1, 1, 512])(outs)
# # outs = K.pool2d(outs, (7,7), pool_mode='avg')
outs = AveragePooling2D((14, 14), name='avg_pool')(convLSTM_outs)
outs = Flatten()(outs)
#
attri_outs = Dense(1000,
kernel_initializer='normal',
activation='sigmoid',
name='attri',
trainable=trainable)(outs)
#
cate_outs = Dense(cate_num,
kernel_initializer='normal',
activation='softmax',
name='cate',
trainable=trainable)(outs)
#
type_outs = Dense(type_num,
kernel_initializer='normal',
activation='softmax',
name='type',
trainable=trainable)(outs)
# land_outs = Dense(196, kernel_initializer='normal', activation='sigmoid', name='land_all', trainable=True)(outs)
# land_outs = Reshape((14, 14, 1))(land_outs)
return [
attri_outs, cate_outs, type_outs, land_outs, up3_land_outs,
up2_land_outs, up1_land_outs
] #
def sam_resnet(x):
#x = [x, x_maps]
# Dilated Convolutional Network
print("Iniciando sam_resnet")
print("Iniciando dcn_resnet...")
dcn = dcn_resnet(input_tensor=x[0]) #Ready!!
aux = K.permute_dimensions(dcn.output, (0, 3, 1, 2)) #Agregado para poner channels_first como en el codigo original.
#conv_feat = Convolution2D(512, 3, 3, border_mode='same', activation='relu')(dcn.output)
# New Version. Input shape = (None, 2048, 30, 40) output shape=(None, 512, 30, 40)
#conv_feat = Conv2D(512,
# (3, 3),
# padding='same',
# activation='relu',
# data_format="channels_first")(aux) # GPU New Version
conv_feat = Conv2D_NCHW(aux,512,
(3, 3),
padding='same',
activation='relu') # CPU New Version NCHW
# Attentive Convolutional LSTM
print("Iniciando att_convlstm...")
att_convlstm = Lambda(repeat, repeat_shape)(conv_feat) #Output shape=(1, 4, 512, 30, 40)
#x = att_convlstm
#l = AttentiveConvLSTM(nb_filters_in=512, nb_filters_out=512, nb_filters_att=512,nb_cols=3, nb_rows=3)
#l(x)
att_convlstm = AttentiveConvLSTM(nb_filters_in=512, #Output shape=(1, 512, 30, 40)
nb_filters_out=512,
nb_filters_att=512,
nb_cols=3,
nb_rows=3)(att_convlstm)
# Learned Prior (1)
#Input shape = (None, 16, 30, 40), output shape = (1, 16, 30, 40)
print("Iniciando LearningPrior 1...")
priors1 = LearningPrior(nb_gaussian=nb_gaussian, init=gaussian_priors_init)(x[1])
#concateneted = merge([att_convlstm, priors1], mode='concat', concat_axis=1)
#print(att_convlstm.shape)
#print(priors1.shape)
#attentive = att_convlstm * 1 #Eliminando un bug
#prior = priors1 * 1 #Eliminando un bug
#print(attentive.shape)
#print(prior.shape)
#concateneted = concatenate([attentive, prior], axis=1) #Nueva version sin BUG
print("Concatenando...")
concateneted = concatenate([att_convlstm, priors1], axis=1) #Version con BUG
#learned_priors1 = AtrousConvolution2D(512, 5, 5, border_mode='same', activation='relu',
# atrous_rate=(4, 4))(concateneted)
#learned_priors1 = Conv2D(512, (5, 5), dilation_rate=(4, 4), activation='relu',
# data_format="channels_first", padding='same')(concateneted) #New version for GPU
learned_priors1 = Conv2D_NCHW(concateneted, 512,
(5, 5),
dilation_rate=(4, 4),
activation='relu',
padding='same') #New version for CPU
# Learned Prior (2)
print("Iniciando LearningPrior 2...")
priors2 = LearningPrior(nb_gaussian=nb_gaussian, init=gaussian_priors_init)(x[1])
#learned_priors1 = learned_priors1 * 1 #Eliminando un bug
#priors2 = priors2 * 1 #Eliminando un bug
#concateneted = merge([learned_priors1, priors2], mode='concat', concat_axis=1)
print("Concatenando...")
concateneted = concatenate([learned_priors1, priors2], axis=1)
#learned_priors2 = AtrousConvolution2D(512, 5, 5, border_mode='same', activation='relu',
# atrous_rate=(4, 4))(concateneted)
#learned_priors2 = Conv2D(512, (5, 5), dilation_rate=(4, 4), activation='relu',
# data_format="channels_first", padding='same')(concateneted) #New version for GPU
learned_priors2 = Conv2D_NCHW(concateneted, 512,
(5, 5),
dilation_rate=(4, 4),
activation='relu',
padding='same') #New version for CPU
# Final Convolutional Layer
print("Final Convolutional Layer")
#outs = Convolution2D(1, 1, 1, border_mode='same', activation='relu')(learned_priors2)
#outs = Conv2D(1, (1, 1), padding='same', data_format="channels_first", activation='relu')(learned_priors2) #New version for GPU output shape=(1, 1, 30, 40)
outs = Conv2D_NCHW(learned_priors2, 1,
(1, 1),
padding='same',
activation='relu') #New version for COU output shape=(1, 1, 30, 40)
#VALIDAR ESTA FUNCION \\\\\\\\\\\\\\\
outs_up = Lambda(upsampling, upsampling_shape)(outs) # Input shape=(1, 1, 30, 40)
#print(outs_up.shape) #(1, 1, 480, 640)
print("Finalizado sam_resnet")
return [outs_up, outs_up, outs_up] #When passing a list as loss, it should have one entry per model outputs