Uncertainty in latent representations of variational autoencoders optimized for visual tasks

Open in new window