Returns a penalty on the mutual information in an InfoGAN model.
tf.contrib.gan.losses.wargs.mutual_information_penalty( structured_generator_inputs, predicted_distributions, weights=1.0, scope=None, loss_collection=tf.GraphKeys.LOSSES, reduction=losses.Reduction.SUM_BY_NONZERO_WEIGHTS, add_summaries=False )
This loss comes from an InfoGAN paper https://arxiv.org/abs/1606.03657.
structured_generator_inputs: A list of Tensors representing the random noise that must have high mutual information with the generator output. List length should match
predicted_distributions: A list of
tfp.distributions.Distributions. Predicted by the recognizer, and used to evaluate the likelihood of the structured noise. List length should match
Tensorwhose rank is either 0, or the same dimensions as
scope: The scope for the operations performed in computing the loss.
loss_collection: collection to which this loss will be added.
tf.compat.v1.losses.Reductionto apply to loss.
add_summaries: Whether or not to add summaries for the loss.
A scalar Tensor representing the mutual information loss.