Module dimdrop.losses.vae_loss
Source code
from keras import backend as K
from keras.metrics import binary_crossentropy
class VAELoss:
"""Custom loss function for a variational autoencoder.
Parameters
----------
in_dim : int
The dimension of the input
z_log_var : tensor
The variance tensor
z_mean : tensor
The mean tensor
"""
__name__ = 'vae_loss'
def __init__(self, in_dim, z_log_var, z_mean):
self.in_dim = in_dim
self.z_log_var = z_log_var
self.z_mean = z_mean
def __call__(self, x, x_decoded_mean):
xent_loss = self.in_dim * \
binary_crossentropy(x, x_decoded_mean)
if self.z_log_var:
kl_loss = - 0.5 * \
K.sum(1 + self.z_log_var - K.square(self.z_mean) -
K.exp(self.z_log_var), axis=-1)
else:
kl_loss = - 0.5 * \
K.sum(1 + 1 - K.square(self.z_mean) - K.exp(1.0), axis=-1)
return K.mean(xent_loss + kl_loss)Classes
class VAELoss (in_dim, z_log_var, z_mean)-
Custom loss function for a variational autoencoder.
Parameters
in_dim:int- The dimension of the input
z_log_var:tensor- The variance tensor
z_mean:tensor- The mean tensor
Source code
class VAELoss: """Custom loss function for a variational autoencoder. Parameters ---------- in_dim : int The dimension of the input z_log_var : tensor The variance tensor z_mean : tensor The mean tensor """ __name__ = 'vae_loss' def __init__(self, in_dim, z_log_var, z_mean): self.in_dim = in_dim self.z_log_var = z_log_var self.z_mean = z_mean def __call__(self, x, x_decoded_mean): xent_loss = self.in_dim * \ binary_crossentropy(x, x_decoded_mean) if self.z_log_var: kl_loss = - 0.5 * \ K.sum(1 + self.z_log_var - K.square(self.z_mean) - K.exp(self.z_log_var), axis=-1) else: kl_loss = - 0.5 * \ K.sum(1 + 1 - K.square(self.z_mean) - K.exp(1.0), axis=-1) return K.mean(xent_loss + kl_loss)