def __init__(self, units, sn=False, **kwargs):
super(vec2img, self).__init__()
self.units_zp = units
self.l1_dense = BaseLayers.Dense(units=units,
use_bias=False,
sn=sn,
**kwargs)
self.l2_norm = BaseLayers.BatchNormalization()
self.l3_activation = BaseLayers.LeakyReLU()
def __init__(self, filters, sn=False, **kwargs):
super(uk_5s1_2D, self).__init__()
self.filters_zp = filters
self.l1_conv2d = BaseLayers.Conv2DTranspose(filters=filters,
kernel_size=[5, 5],
strides=[1, 1],
padding="SAME",
use_bias=False,
sn=sn,
**kwargs)
self.l2_norm = BaseLayers.BatchNormalization()
self.l3_activation = BaseLayers.LeakyReLU()
def __init__(self,filters,sn=False,activation=None,**kwargs):
super(uk_5s2_2D,self).__init__()
self.filters_zp = filters
self.l1_conv2d = BaseLayers.Conv2DTranspose(filters=filters,kernel_size=[5,5],strides=[2,2],padding="SAME",use_bias=False,sn=sn,**kwargs)
self.l2_norm = BaseLayers.BatchNormalization()
if activation == None:
self.l3_activation = tf.keras.layers.Activation("linear")
elif activation.lower()=="relu":
self.l3_activation = BaseLayers.ReLU()
elif activation.lower()=="leaky_relu":
self.l3_activation = BaseLayers.LeakyReLU()
elif activation.lower()=="sigmoid":
self.l3_activation = tf.keras.layers.Activation("sigmoid")
elif activation.lower()=="tanh":
self.l3_activation = tf.keras.layers.Activation("tanh")
else:
raise ValueError("Un supported activation"+activation)