When used as a surrogate objective for maximum likelihood estimation in latent variable models, the evidence lower bound (ELBO) produces state-of-the-art results.
Learning in models with discrete latent variables is challenging due to high variance gradient estimators.
The dominant sequence transduction models are based on complex recurrent or convolutional neural networks in an encoder-decoder configuration.
We prove that, since the data instances with larger gradients play a more important role in the computation of information gain, GOSS can obtain quite accurate estimation of the information gain with a much smaller data size.
Generative Adversarial Networks (GANs) excel at creating realistic images with complex models for which maximum likelihood is infeasible.
#4 best model for
Image Generation
on LSUN Bedroom 256 x 256
Understanding why a model makes a certain prediction can be as crucial as the prediction's accuracy in many applications.
Generative Adversarial Networks (GANs) are powerful generative models, but suffer from training instability.
#2 best model for
Image Generation
on CAT 256x256
We present a practical way of introducing convolutional structure into Gaussian processes, making them more suited to high-dimensional inputs like images.
Existing approaches to inference in DGP models assume approximate posteriors that force independence between the layers, and do not work well in practice.
Unsupervised image-to-image translation aims at learning a joint distribution of images in different domains by using images from the marginal distributions in individual domains.
#2 best model for
Multimodal Unsupervised Image-To-Image Translation
on Cats-and-Dogs
DOMAIN ADAPTATION MULTIMODAL UNSUPERVISED IMAGE-TO-IMAGE TRANSLATION UNSUPERVISED IMAGE-TO-IMAGE TRANSLATION