def __init__(self,
image_size,
learning_rate=2e-5,
batch_size=1,
ngf=64,
units=4096
):
"""
Args:
input_size：list [H, W, C]
batch_size: integer, batch size
learning_rate: float, initial learning rate for Adam
ngf: number of gen filters in first conv layer
"""
self.learning_rate = learning_rate
self.input_shape = [int(batch_size / 4), image_size[0], image_size[1], image_size[2]]
self.ones = tf.ones(self.input_shape, name="ones")
self.tenaor_name = {}
self.EC_S = VEncoder('EC_S', ngf=ngf, units=units, keep_prob=0.85)
self.DC_S = VDecoder('DC_S', ngf=ngf, output_channl=2, units=units)
self.G_M = Unet('G_M', ngf=ngf / 2, keep_prob=0.9, output_channl=2)
self.D_S = Discriminator('D_S', ngf=ngf, keep_prob=0.85)
self.FD_Z = FeatureDiscriminator('FD_Z', ngf=ngf)
def __init__(self,
image_size,
learning_rate=2e-5,
batch_size=1,
ngf=64,
):
"""
Args:
input_size：list [H, W, C]
batch_size: integer, batch size
learning_rate: float, initial learning rate for Adam
ngf: number of gen filters in first conv layer
"""
self.learning_rate = learning_rate
self.input_shape = [int(batch_size / 4), image_size[0], image_size[1], image_size[2]]
self.code_shape = [int(batch_size / 4), int(image_size[0] / 4), int(image_size[1] / 4), 4]
self.ones = tf.ones(self.input_shape, name="ones")
self.ones_code = tf.ones(self.code_shape, name="ones_code")
self.image_list = {}
self.prob_list = {}
self.code_list = {}
self.judge_list = {}
self.tenaor_name = {}
self.EC_L = Encoder('EC_L', ngf=ngf)
self.DC_L = Decoder('DC_L', ngf=ngf, output_channl=5)
self.EC_R = Encoder('EC_R', ngf=ngf)
self.EC_M = Encoder('EC_M', ngf=ngf)
self.DC_M = Decoder('DC_M', ngf=ngf)
self.D_M = Discriminator('D_M', ngf=ngf)
self.FD_R = FeatureDiscriminator('FD_R', ngf=ngf)