Spotlight
Adaptive Density Estimation for Generative Models
Thomas Lucas · Konstantin Shmelkov · Karteek Alahari · Cordelia Schmid · Jakob Verbeek

Tue Dec 10th 05:05 -- 05:10 PM @ West Exhibition Hall C + B3

Unsupervised learning of generative models has seen tremendous progress over recent years, in particular due to generative adversarial networks (GANs), variational autoencoders, and flow-based models. GANs have dramatically improved sample quality, but suffer from two drawbacks: (i) they mode-drop, \ie, do not cover the full support of the train data, and (ii) they do not allow for likelihood evaluations on held-out data. In contrast likelihood-based training encourages models to cover the full support of the train data, but yields poorer samples. These mutual shortcomings can in principle be addressed by training generative latent variable models in a hybrid adversarial-likelihood manner. However, we show that commonly made parametric assumptions create a conflict between them, making successful hybrid models non trivial. As a solution, we propose the use of deep invertible transformations in the latent variable decoder. This approach allows for likelihood computations in image space, is more efficient than fully invertible models, and can take full advantage of adversarial training. We show that our model significantly improves over existing hybrid models: offering GAN-like samples, IS and FID scores that are competitive with fully adversarial models and improved likelihood scores.

Author Information

Thomas Lucas (Inria)

I am a Ph.d student at Inria Grenoble in team THOTH (formerly team LEAR), dealing with computer vision and machine learning.

Konstantin Shmelkov (Huawei)
Karteek Alahari (Inria)
Cordelia Schmid (Inria / Google)
Jakob Verbeek (INRIA)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors