def _custom_discriminator(input_net, class_num):
with tf.compat.v1.variable_scope('discriminator'):
hidden_src = layers.discriminator_input_hidden(
input_net, scope='discriminator_input_hidden_source')
output_src = layers.discriminator_output_source(hidden_src)
output_src = tf.compat.v1.layers.flatten(output_src)
output_src = tf.reduce_mean(input_tensor=output_src, axis=1)
hidden_cls = layers.discriminator_input_hidden(input_net,
trainable=False)
output_cls = layers.discriminator_output_class(hidden_cls,
class_num,
trainable=False)
return output_src, output_cls
def discriminator(input_net, class_num):
"""Discriminator Module.
Piece everything together and reshape the output source tensor
PyTorch Version:
https://github.com/yunjey/StarGAN/blob/fbdb6a6ce2a4a92e1dc034faec765e0dbe4b8164/model.py#L63
Notes:
The PyTorch Version run the reduce_mean operation later in their solver:
https://github.com/yunjey/StarGAN/blob/fbdb6a6ce2a4a92e1dc034faec765e0dbe4b8164/solver.py#L245
Args:
input_net: Tensor of shape (batch_size, h, w, c) as batch of images.
class_num: (int) number of domain to be predicted
Returns:
output_src: Tensor of shape (batch_size) where each value is a logit
representing whether the image is real of fake.
output_cls: Tensor of shape (batch_size, class_um) where each value is a
logit representing whether the image is in the associated domain.
"""
with tf.variable_scope('discriminator'):
hidden = layers.discriminator_input_hidden(input_net)
output_src = layers.discriminator_output_source(hidden)
output_src = tf.layers.flatten(output_src)
output_src = tf.reduce_mean(input_tensor=output_src, axis=1)
output_cls = layers.discriminator_output_class(hidden, class_num)
return output_src, output_cls
def __call__(self, input_net, class_num):
with tf.compat.v1.variable_scope('discriminator'):
hidden_src = layers.discriminator_input_hidden(
input_net, scope='discriminator_input_hidden_source')
output_src = layers.discriminator_output_source(hidden_src)
output_src = tf.compat.v1.layers.flatten(output_src)
output_src = tf.reduce_mean(input_tensor=output_src, axis=1)
output_cls = self.keras_model((input_net + 1.0) / 2.0)
return output_src, output_cls
def test_discriminator_input_hidden(self):
n = 2
h = 128
w = 128
c = 3
input_tensor = tf.random.uniform((n, h, w, c))
output_tensor = layers.discriminator_input_hidden(input_tensor)
with self.cached_session() as sess:
sess.run(tf.global_variables_initializer())
output = sess.run(output_tensor)
self.assertTupleEqual((n, 2, 2, 2048), output.shape)