def discriminator(self, image, t_text_embedding, reuse=False): if reuse: tf.get_variable_scope().reuse_variables() self.w_d = [w for w in tf.global_variables()] h0 = ops.lrelu(ops.conv2d(image, self.options['df_dim'], name = 'd_h0_conv')) #32 h1 = ops.lrelu( self.d_bn1(ops.conv2d(h0, self.options['df_dim']*2, name = 'd_h1_conv'))) #16 h2 = ops.lrelu( self.d_bn2(ops.conv2d(h1, self.options['df_dim']*4, name = 'd_h2_conv'))) #8 h3 = ops.lrelu( self.d_bn3(ops.conv2d(h2, self.options['df_dim']*8, name = 'd_h3_conv'))) #4 # ADD TEXT EMBEDDING TO THE NETWORK reduced_text_embeddings = ops.lrelu(ops.linear(t_text_embedding, self.options['t_dim'], 'd_embedding')) reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,1) reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,2) tiled_embeddings = tf.tile(reduced_text_embeddings, [1,4,4,1], name='tiled_embeddings') h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat') h3_new = ops.lrelu( self.d_bn4(ops.conv2d(h3_concat, self.options['df_dim']*8, 1,1,1,1, name = 'd_h3_conv_new'))) #4 h4 = ops.linear(tf.reshape(h3_new, [self.options['batch_size'], -1]), 1, 'd_h3_lin') #return tf.nn.sigmoid(h4), h4 clip_updates = [w.assign(tf.clip_by_value(w, -1e-2, 1e-2)) for w in self.w_d] if self.option['gan_type'] == 1: return h4, h4, clip_updates else return tf.nn.sigmoid(h4), h4, clip_updates
def discriminator(self, image, t_text_embedding, reuse=False):
with tf.variable_scope('discriminator'):
if reuse:
tf.get_variable_scope().reuse_variables()
h0 = ops.lrelu(ops.conv2d(image, self.options['df_dim'],
name = 'd_h0_conv')) #32
h1 = ops.lrelu(self.d_bn1(ops.conv2d(h0, self.options['df_dim']*2,
name='d_h1_conv'))) #16
h2 = ops.lrelu(self.d_bn2(ops.conv2d(h1, self.options['df_dim']*4,
name='d_h2_conv'))) #8
h3 = ops.lrelu(self.d_bn3(ops.conv2d(h2, self.options['df_dim']*8,
name='d_h3_conv'))) #4
# ADD TEXT EMBEDDING TO THE NETWORK
reduced_text_embeddings =\
ops.lrelu(ops.linear(t_text_embedding,
self.options['t_dim'], 'd_embedding'))
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,1)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,2)
tiled_embeddings =\
tf.tile(reduced_text_embeddings, [1,4,4,1], name='tiled_embeddings')
h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
h3_new =\
ops.lrelu( self.d_bn4(ops.conv2d(h3_concat, self.options['df_dim']*8, 1,
1, 1, 1, name='d_h3_conv_new'))) #4
h4 = ops.linear(tf.reshape(h3_new, [self.options['batch_size'], -1]),
1, 'd_h3_lin')
return tf.nn.sigmoid(h4), h4
def downsampling(self, image, reuse = False):
if reuse:
tf.get_variable_scope().reuse_variables()
# Downsample to 32*32*256
h0 = ops.lrelu( ops.conv2d(image, 256,
name = 'down_h0_conv')) #32
# Downsample to 16*16*512
h1 = ops.lrelu( self.down_bn1(ops.conv2d(h0, 512,
name = 'down_h1_conv'))) #32
return h1
def discriminator(self, image, t_text_embedding):
with tf.variable_scope("discriminator", reuse=tf.AUTO_REUSE):
h0 = ops.lrelu(
ops.conv2d(image, self.options['df_dim'],
name='d_h0_conv')) #32
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d(h0,
self.options['df_dim'] * 2,
name='d_h1_conv'))) #16
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d(h1,
self.options['df_dim'] * 4,
name='d_h2_conv'))) #8
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d(h2,
self.options['df_dim'] * 8,
name='d_h3_conv'))) #4
# ADD TEXT EMBEDDING TO THE NETWORK
reduced_text_embeddings = ops.lrelu(
ops.linear(t_text_embedding, self.options['t_dim'],
'd_embedding'))
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
1)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
2)
tiled_embeddings = tf.tile(reduced_text_embeddings, [1, 4, 4, 1],
name='tiled_embeddings')
h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
h3_new = ops.lrelu(
self.d_bn4(
ops.conv2d(h3_concat,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='d_h3_conv_new'))) #4
h4 = ops.linear(
tf.reshape(h3_new, [self.options['batch_size'], -1]), 1,
'd_h3_lin')
# Auxiliary classifier
class_logit = ops.linear(
tf.reshape(h3_new, [self.options['batch_size'], -1]),
self.options['num_class'], 'd_image_classifier')
return h4, h4, class_logit
def discriminator(self, image, t_text_embedding, reuse=False):
with tf.variable_scope("discriminator", reuse=tf.AUTO_REUSE):
h0 = ops.lrelu(
ops.conv2d(image, self.options['df_dim'],
name='d_h0_conv')) #32
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d(h0,
self.options['df_dim'] * 2,
name='d_h1_conv'))) #16
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d(h1,
self.options['df_dim'] * 4,
name='d_h2_conv'))) #8
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d(h2,
self.options['df_dim'] * 8,
name='d_h3_conv'))) #4
h3_new = ops.lrelu(
self.d_bn4(
ops.conv2d(h3,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='d_h3_conv_new'))) #4
h3_new = tf.reshape(h3_new, [self.options['batch_size'], -1])
image_embedding = ops.linear(h3_new, self.options['t_dim'],
'd_h3_embedding')
# Embedding matrix of condition
reduced_text_embeddings = ops.linear(t_text_embedding,
self.options['t_dim'],
'd_embedding')
# Scalar output function
h4 = ops.linear(image_embedding, 1, 'd_scalar_output')
discriminator_output_logit = tf.reduce_sum(tf.multiply(
reduced_text_embeddings, image_embedding),
1,
keepdims=True) + h4
return tf.nn.sigmoid(
discriminator_output_logit), discriminator_output_logit
def generator(self, t_z, t_text_embedding):
s = self.options['image_size']
s2, s4, s8, s16 = int(s/2), int(s/4), int(s/8), int(s/16)
reduced_text_embedding = ops.lrelu( ops.linear(t_text_embedding, self.options['t_dim'], 'g_embedding') )
tf.summary.tensor_summary("Reduced voice embedding", reduced_text_embedding)
#z_concat = tf.concat(1, [t_z, reduced_text_embedding])
z_concat = t_text_embedding
z_ = ops.linear(z_concat, self.options['gf_dim']*8*s16*s16, 'g_h0_lin')
h0 = tf.reshape(z_, [-1, s16, s16, self.options['gf_dim'] * 8])
h0 = tf.nn.relu(self.g_bn0(h0))
h1 = ops.deconv2d(h0, [self.options['batch_size'], s8, s8, self.options['gf_dim']*4], name='g_h1')
h1 = tf.nn.relu(self.g_bn1(h1))
h2 = ops.deconv2d(h1, [self.options['batch_size'], s4, s4, self.options['gf_dim']*2], name='g_h2')
h2 = tf.nn.relu(self.g_bn2(h2))
h3 = ops.deconv2d(h2, [self.options['batch_size'], s2, s2, self.options['gf_dim']*1], name='g_h3')
h3 = tf.nn.relu(self.g_bn3(h3))
# Classify class
h3_new = ops.lrelu(self.g_bn4(ops.conv2d(h3, self.options['df_dim'], 1, 1, 1, 1, name="g_conv")))
h3_new = tf.reshape(h3_new, [self.options['batch_size'], -1])
class_logit = ops.linear(h3_new, self.options['num_class'], 'g_h3_embedding')
h4 = ops.deconv2d(h3, [self.options['batch_size'], s, s, 3], name='g_h4')
return (tf.tanh(h4)/2. + 0.5), class_logit
def discriminator(self, image, reuse=False):
if reuse:
tf.get_variable_scope().reuse_variables()
h0 = ops.lrelu( ops.conv2d(image, 64,
name = 'd_h0_conv')) #32
h1 = ops.lrelu( self.d_bn1(ops.conv2d(h0, 128,
name = 'd_h1_conv'))) #32
h2 = ops.lrelu( self.d_bn2(ops.conv2d(h1, 256,
name = 'd_h2_conv'))) #32
h3 = ops.lrelu( self.d_bn3(ops.conv2d(h2, 512,
name = 'd_h3_conv'))) #32
h4 = ops.linear(tf.reshape(h3, [self.options['batch_size'], -1]), 1, 'd_h3_lin')
return tf.nn.sigmoid(h4), h4
def classifier(self, image, n_classes, t_training, reuse=False): if reuse: tf.get_variable_scope().reuse_variables() h0 = ops.lrelu(ops.conv2d(image, self.options['ef_dim'] * 8, name='e_h0_conv')) # 64 h1 = ops.lrelu(slim.batch_norm(ops.conv2d(h0, self.options['ef_dim'] * 8, name='e_h1_conv'), reuse=reuse, is_training=t_training, scope='e_bn1')) # 32 h2 = ops.lrelu(slim.batch_norm(ops.conv2d(h1, self.options['ef_dim'] * 6, name='e_h2_conv'), reuse=reuse, is_training=t_training, scope='e_bn2')) # 16 h3 = ops.lrelu(slim.batch_norm(ops.conv2d(h2, self.options['ef_dim'] * 4, name='e_h3_conv'), reuse=reuse, is_training=t_training, scope='e_bn3')) # 8 h4 = ops.lrelu(slim.batch_norm(ops.conv2d(h3, self.options['ef_dim'] * 2, name='e_h4_conv'), reuse=reuse, is_training=t_training, scope='e_bn4')) # 8 h4_shape = h4.get_shape().as_list() h4_flat = tf.contrib.layers.flatten(h4) ''' h4_flat = tf.squeeze(tf.nn.avg_pool(h4, ksize=[1, h4_shape[1], h4_shape[2], 1], strides=[1, h4_shape[1], h4_shape[2], 1], padding='SAME', name='global_avg_pool')) ''' fc1 = tf.nn.relu(ops.linear(h4_flat, 1024, 'fl_01')) fc2 = ops.linear(fc1, n_classes, 'fl_02') return fc2
def discriminator(self, image, t_text_embedding, reuse=False):
with tf.variable_scope(tf.get_variable_scope(), reuse=reuse):
# 64
print ('check')
print (image.shape)
h0 = ops.lrelu(ops.conv2d(image, self.options['df_dim'] * 4, d_w = 16, name='d_h0_conv')) # 32
print (h0.shape)
# h1 = ops.lrelu( self.d_bn1(ops.conv2d(h0, self.options['df_dim']*2, name = 'd_h1_conv'))) #16
# print (h1.shape)
# h2 = ops.lrelu( self.d_bn2(ops.conv2d(h1, self.options['df_dim']*4, name = 'd_h2_conv'))) #8
# print (h2.shape)
# h3 = ops.lrelu( self.d_bn3(ops.conv2d(h2, self.options['df_dim']*8, name = 'd_h3_conv'))) #4
# print (h3.shape)
h3 = h0
# ADD TEXT EMBEDDING TO THE NETWORK
# 256
print ('check')
print (t_text_embedding.shape)
reduced_text_embeddings = ops.lrelu(ops.linear(t_text_embedding, self.options['t_dim'], 'd_embedding'))
print (reduced_text_embeddings.shape)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings, 1)
print (reduced_text_embeddings.shape)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings, 2)
print (reduced_text_embeddings.shape)
tiled_embeddings = tf.tile(reduced_text_embeddings, [1, 4, 4, 1], name='tiled_embeddings')
print (tiled_embeddings.shape)
print (h3)
print (h3.shape)
print (tiled_embeddings)
print (tiled_embeddings.shape)
h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
#error:
# ValueError: Dimensions must be equal, but are 176 and 1 for 'd_h3_conv_new/Conv2D' (op: 'Conv2D') with input shapes: [64,4,4,176], [1,1,1,88].
tmp = ops.conv2d(h3_concat, self.options['df_dim'] * 4, 1, 1, 1, 1, name='d_h3_conv_new')
h3_new = ops.lrelu(self.d_bn4(tmp)) # 4
h4 = ops.linear(tf.reshape(h3_new, [self.options['batch_size'], -1]), 1, 'd_h3_lin')
return tf.nn.sigmoid(h4), h4
def discriminator(self, image, t_text_embedding):
update_collection = tf.GraphKeys.UPDATE_OPS
with tf.variable_scope("discriminator", reuse=tf.AUTO_REUSE):
h0 = ops.lrelu(
ops.conv2d_sn(image,
self.options['df_dim'],
spectral_normed=True,
update_collection=update_collection,
name='d_h0_conv')) #32
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d_sn(h0,
self.options['df_dim'] * 2,
spectral_normed=True,
update_collection=update_collection,
name='d_h1_conv'))) #16
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d_sn(h1,
self.options['df_dim'] * 4,
spectral_normed=True,
update_collection=update_collection,
name='d_h2_conv'))) #8
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d_sn(h2,
self.options['df_dim'] * 8,
spectral_normed=True,
update_collection=update_collection,
name='d_h3_conv'))) #4
# ADD TEXT EMBEDDING TO THE NETWORK
reduced_text_embeddings = ops.lrelu(
ops.linear(t_text_embedding, self.options['t_dim'],
'd_embedding'))
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
1)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
2)
tiled_embeddings = tf.tile(reduced_text_embeddings, [1, 4, 4, 1],
name='tiled_embeddings')
h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
h3_new = ops.lrelu(
self.d_bn4(
ops.conv2d(h3_concat,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='d_h3_conv_new'))) #4
h4 = ops.linear(
tf.reshape(h3_new, [self.options['batch_size'], -1]), 1,
'd_h3_lin')
return h4, h4
def encoder(self, image, t_training, reuse = False) : if reuse : tf.get_variable_scope().reuse_variables() h0 = ops.lrelu(ops.conv2d(image, self.options['ef_dim'] * 8, name = 'e_h0_conv')) # 64 h1 = ops.lrelu(slim.batch_norm(ops.conv2d(h0, self.options['ef_dim'] * 8, name = 'e_h1_conv'), reuse=reuse, is_training = t_training, scope = 'e_bn1')) # 32 h2 = ops.lrelu(slim.batch_norm(ops.conv2d(h1, self.options['ef_dim'] * 6, name = 'e_h2_conv'), reuse=reuse, is_training = t_training, scope = 'e_bn2')) # 16 h3 = ops.lrelu(slim.batch_norm(ops.conv2d(h2, self.options['ef_dim'] * 4, name = 'e_h3_conv'), reuse=reuse, is_training = t_training, scope = 'e_bn3')) # 8 h4 = ops.lrelu(slim.batch_norm(ops.conv2d(h3, self.options['ef_dim'] * 2, name = 'e_h4_conv'), reuse=reuse, is_training = t_training, scope = 'e_bn4')) # 8 h4_shape = h4.get_shape().as_list() #h4_flat = tf.contrib.layers.flatten(h4) #h5 = ops.linear(h4_flat, 1024, 'fl_e_01') #h6 = ops.linear(h5, n_classes, 'fl_e_02') return h4, h4_shape
def discriminator(self, image, t_text_embedding, reuse=False): if reuse: tf.get_variable_scope().reuse_variables() h0 = ops.lrelu(ops.conv2d(image, self.options['df_dim'], name = 'd_h0_conv')) #32 h1 = ops.lrelu( self.d_bn1(ops.conv2d(h0, self.options['df_dim']*2, name = 'd_h1_conv'))) #16 h2 = ops.lrelu( self.d_bn2(ops.conv2d(h1, self.options['df_dim']*4, name = 'd_h2_conv'))) #8 h3 = ops.lrelu( self.d_bn3(ops.conv2d(h2, self.options['df_dim']*8, name = 'd_h3_conv'))) #4 # ADD TEXT EMBEDDING TO THE NETWORK reduced_text_embeddings = ops.lrelu(ops.linear(t_text_embedding, self.options['t_dim'], 'd_embedding')) reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,1) reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,2) tiled_embeddings = tf.tile(reduced_text_embeddings, [1,4,4,1], name='tiled_embeddings') h3_concat = tf.concat( 3, [h3, tiled_embeddings], name='h3_concat') h3_new = ops.lrelu( self.d_bn4(ops.conv2d(h3_concat, self.options['df_dim']*8, 1,1,1,1, name = 'd_h3_conv_new'))) #4 h4 = ops.linear(tf.reshape(h3_new, [self.options['batch_size'], -1]), 1, 'd_h3_lin') return tf.nn.sigmoid(h4), h4
def image_encoder(self, image):
with tf.variable_scope("image_vae", reuse=tf.AUTO_REUSE):
h0 = ops.lrelu(
ops.conv2d(image, self.options['df_dim'],
name='v_h0_conv')) #32
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d(h0,
self.options['df_dim'] * 2,
name='v_h1_conv'))) #16
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d(h1,
self.options['df_dim'] * 4,
name='v_h2_conv'))) #8
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d(h2,
self.options['df_dim'] * 8,
name='v_h3_conv'))) #4
h3_new = ops.lrelu(
self.d_bn4(
ops.conv2d(h3,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='v_h3_conv_new')))
hidden = tf.reshape(h3_new, [self.options['batch_size'], -1])
mean = ops.linear(hidden, self.options['t_dim'], 'v_enc_mean')
std = 1e-6 + tf.nn.softplus(
ops.linear(hidden, self.options['t_dim'], 'v_enc_std'))
code = mean + std * tf.random_normal(
mean.get_shape().as_list(), 0, 1, dtype=tf.float32)
return code, mean, std
def residuel_block(self, input_tensor, reuse = False):
if reuse:
tf.get_variable_scope().reuse_variables()
#[To do] it seemed that after identity mapping, there should be a relu
h0 = tf.nn.relu(self.r_bn0(ops.conv2d(input_tensor, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h0')))
h1 = self.r_bn1(ops.conv2d(h0, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h1'))
h2 = input_tensor + h1
h3 = tf.nn.relu(self.r_bn3(ops.conv2d(h2, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h3')))
h4 = self.r_bn4(ops.conv2d(h3, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h4'))
h5 = h4 + h2
h6 = tf.nn.relu(self.r_bn6(ops.conv2d(h5, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h6')))
h7 = self.r_bn7(ops.conv2d(h6, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h7'))
h8 = h7 + h5
h9 = tf.nn.relu(self.r_bn9(ops.conv2d(h8, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h9')))
h10 = self.r_bn10(ops.conv2d(h9, 512,
k_h=3, k_w=3, d_h=1, d_w=1,
name = 'res_h10'))
h8 = h10 + h8
return h8
def discriminator(self, image, t_text_embedding, reuse=False):
if reuse:
with tf.variable_scope(tf.get_variable_scope(), reuse=True):
h0 = ops.lrelu(
ops.conv2d(image, self.options['df_dim'],
name='d_h0_conv')) #32,48
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d(h0,
self.options['df_dim'] * 2,
name='d_h1_conv'))) #16,24
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d(h1,
self.options['df_dim'] * 4,
name='d_h2_conv'))) #8,12
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d(h2,
self.options['df_dim'] * 8,
name='d_h3_conv'))) #4,6
# ADD TEXT EMBEDDING TO THE NETWORK
reduced_text_embeddings = ops.lrelu(
ops.linear(t_text_embedding, self.options['t_dim'],
'd_embedding'))
reduced_text_embeddings = tf.expand_dims(
reduced_text_embeddings, 1)
reduced_text_embeddings = tf.expand_dims(
reduced_text_embeddings, 2)
tiled_embeddings = tf.tile(reduced_text_embeddings,
[1, 4, 4, 1],
name='tiled_embeddings')
# h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
h3_concat = tf.concat(values=[h3, tiled_embeddings],
axis=3,
name='h3_concat')
h3_new = ops.lrelu(
self.d_bn4(
ops.conv2d(h3_concat,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='d_h3_conv_new'))) #4
h4 = ops.linear(
tf.reshape(h3_new, [self.options['batch_size'], -1]), 1,
'd_h3_lin')
# return tf.nn.sigmoid(h4), h4
else:
h0 = ops.lrelu(
ops.conv2d(image, self.options['df_dim'],
name='d_h0_conv')) #32,48
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d(h0,
self.options['df_dim'] * 2,
name='d_h1_conv'))) #16,24
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d(h1,
self.options['df_dim'] * 4,
name='d_h2_conv'))) #8,12
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d(h2,
self.options['df_dim'] * 8,
name='d_h3_conv'))) #4,6
# ADD TEXT EMBEDDING TO THE NETWORK
reduced_text_embeddings = ops.lrelu(
ops.linear(t_text_embedding, self.options['t_dim'],
'd_embedding'))
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
1)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
2)
tiled_embeddings = tf.tile(reduced_text_embeddings, [1, 4, 4, 1],
name='tiled_embeddings')
# h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
h3_concat = tf.concat(values=[h3, tiled_embeddings],
axis=3,
name='h3_concat')
h3_new = ops.lrelu(
self.d_bn4(
ops.conv2d(h3_concat,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='d_h3_conv_new'))) #4
h4 = ops.linear(
tf.reshape(h3_new, [self.options['batch_size'], -1]), 1,
'd_h3_lin')
# return tf.nn.sigmoid(h4), h4
return tf.nn.sigmoid(h4), h4
# if reuse:
# tf.get_variable_scope().reuse_variables()
# h0 = ops.lrelu(ops.conv2d(image, self.options['df_dim'], name = 'd_h0_conv')) #32,48
# h1 = ops.lrelu( self.d_bn1(ops.conv2d(h0, self.options['df_dim']*2, name = 'd_h1_conv'))) #16,24
# h2 = ops.lrelu( self.d_bn2(ops.conv2d(h1, self.options['df_dim']*4, name = 'd_h2_conv'))) #8,12
# h3 = ops.lrelu( self.d_bn3(ops.conv2d(h2, self.options['df_dim']*8, name = 'd_h3_conv'))) #4,6
# # ADD TEXT EMBEDDING TO THE NETWORK
# reduced_text_embeddings = ops.lrelu(ops.linear(t_text_embedding, self.options['t_dim'], 'd_embedding'))
# reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,1)
# reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,2)
# tiled_embeddings = tf.tile(reduced_text_embeddings, [1,6,6,1], name='tiled_embeddings')
# # h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
# h3_concat = tf.concat(values=[h3, tiled_embeddings], axis=3, name='h3_concat')
# h3_new = ops.lrelu( self.d_bn4(ops.conv2d(h3_concat, self.options['df_dim']*8, 1,1,1,1, name = 'd_h3_conv_new'))) #4
# h4 = ops.linear(tf.reshape(h3_new, [self.options['batch_size'], -1]), 1, 'd_h3_lin')
# return tf.nn.sigmoid(h4), h4
def discriminator(self, image, t_text_embedding):
with tf.variable_scope("discriminator", reuse=tf.AUTO_REUSE):
if self.options['vgg']:
h0 = ops.lrelu(
ops.conv2d(image, 3, stride=1, name='d_h0a_conv'))
h0 = ops.lrelu(ops.conv2d(h0, 3, stride=2, name='d_h0_conv'))
h1 = ops.lrelu(
self.d_bn1a(
ops.conv2d(h0,
self.options['df_dim'] * 2,
stride=1,
name='d_h1a_conv')))
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d(h0,
self.options['df_dim'] * 2,
stride=2,
name='d_h1_conv')))
h2 = ops.lrelu(
self.d_bn2a(
ops.conv2d(h1,
self.options['df_dim'] * 4,
stride=1,
name='d_h2a_conv')))
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d(h2,
self.options['df_dim'] * 4,
stride=2,
name='d_h2_conv')))
h3 = ops.lrelu(
self.d_bn3a(
ops.conv2d(h2,
self.options['df_dim'] * 8,
stride=1,
name='d_h3a_conv')))
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d(h3,
self.options['df_dim'] * 8,
stride=2,
name='d_h3_conv')))
else:
if self.options['extra_64']:
image = ops.lrelu(
ops.conv2d(image, 3, stride=1, name='d_h0a_conv'))
h0 = ops.lrelu(
ops.conv2d(image, self.options['df_dim'],
name='d_h0_conv')) # 32
if self.options['extra_32']:
h0 = ops.lrelu(
self.d_bn1a(
ops.conv2d(h0,
self.options['df_dim'],
stride=1,
name='d_h1a_conv')))
h1 = ops.lrelu(
self.d_bn1(
ops.conv2d(h0,
self.options['df_dim'] * 2,
name='d_h1_conv'))) # 16
h2 = ops.lrelu(
self.d_bn2(
ops.conv2d(h1,
self.options['df_dim'] * 4,
name='d_h2_conv'))) # 8
h3 = ops.lrelu(
self.d_bn3(
ops.conv2d(h2,
self.options['df_dim'] * 8,
name='d_h3_conv'))) # 4
# ADD TEXT EMBEDDING TO THE NETWORK
reduced_text_embeddings = ops.lrelu(
ops.linear(t_text_embedding, self.options['t_dim'],
'd_embedding'))
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
1)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings,
2)
tiled_embeddings = tf.tile(reduced_text_embeddings, [1, 4, 4, 1],
name='tiled_embeddings')
h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
h3_new = ops.lrelu(
self.d_bn4(
ops.conv2d(h3_concat,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='d_h3_conv_new'))) # 4
h4 = ops.linear(
tf.reshape(h3_new, [self.options['batch_size'], -1]), 1,
'd_h3_lin')
return tf.nn.sigmoid(h4), h4
def discriminator(self,
image,
t_text_embedding,
n_classes,
t_training,
reuse=False):
if reuse:
tf.get_variable_scope().reuse_variables()
h0 = ops.lrelu(
ops.conv2d(image, self.options['df_dim'], name='d_h0_conv')) # 64
h1 = ops.lrelu(
slim.batch_norm(ops.conv2d(h0,
self.options['df_dim'] * 2,
name='d_h1_conv'),
reuse=reuse,
is_training=t_training,
scope='d_bn1')) # 32
h2 = ops.lrelu(
slim.batch_norm(ops.conv2d(h1,
self.options['df_dim'] * 4,
name='d_h2_conv'),
reuse=reuse,
is_training=t_training,
scope='d_bn2')) # 16
h3 = ops.lrelu(
slim.batch_norm(ops.conv2d(h2,
self.options['df_dim'] * 8,
name='d_h3_conv'),
reuse=reuse,
is_training=t_training,
scope='d_bn3')) # 8
h3_shape = h3.get_shape().as_list()
# ADD TEXT EMBEDDING TO THE NETWORK
reduced_text_embeddings = ops.lrelu(
ops.linear(t_text_embedding, self.options['t_dim'], 'd_embedding'))
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings, 1)
reduced_text_embeddings = tf.expand_dims(reduced_text_embeddings, 2)
tiled_embeddings = tf.tile(reduced_text_embeddings,
[1, h3_shape[1], h3_shape[1], 1],
name='tiled_embeddings')
h3_concat = tf.concat([h3, tiled_embeddings], 3, name='h3_concat')
h3_new = ops.lrelu(
slim.batch_norm(ops.conv2d(h3_concat,
self.options['df_dim'] * 8,
1,
1,
1,
1,
name='d_h3_conv_new'),
reuse=reuse,
is_training=t_training,
scope='d_bn4')) # 4
h3_flat = tf.reshape(h3_new, [self.options['batch_size'], -1])
h4 = ops.linear(h3_flat, 1, 'd_h4_lin_rw')
h4_aux = ops.linear(h3_flat, n_classes, 'd_h4_lin_ac')
return tf.nn.sigmoid(h4), h4, tf.nn.sigmoid(h4_aux), h4_aux
