Primal-Dual Wasserstein GAN
By Mevlana Gemici, Zeynep Akata, Max Welling. 2018
14 citations recorded by the source corpus.
We introduce Primal-Dual Wasserstein GAN, a new learning algorithm for building latent variable models of the data distribution based on the primal and the dual formulations of the optimal transport (OT) problem. We utilize the primal formulation to learn a flexible inference mechanism and to create an optimal approximate coupling between the data distribution and the generative model. In order to learn the generative model, we use the dual formulation and train the decoder adversarially through a critic network that is regularized by the approximate coupling obtained from the primal. Unlike previous methods that violate various properties of the optimal critic, we regularize the norm and the direction of the gradients of the critic function. Our model shares many of the desirable properties of auto-encoding models in terms of mode coverage and latent structure, while avoiding their undesirable averaging properties, e.g. their inability to capture sharp visual features when modeling real images. We compare our algorithm with several other generative modeling techniques that utilize Wasserstein distances on Frechet Inception Distance (FID) and Inception Scores (IS).
Subfields: Ocean Engineering, Radiology, Nuclear Medicine and Imaging
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Ocean Engineering
In Radiology, Nuclear Medicine and Imaging
Citations in the corpus
- Sorting out Lipschitz function approximation (2018, methodology): “1 We note that optimal surfaces under the dual Wasserstein objective have a gradient norm of 1 almost everywhere (Corollary 1 in Gemici et al. [16]).”
- A Domain Agnostic Measure for Monitoring and Evaluating GANs (2018, methodology): “A similar approach proposed by [7] uses the dual formulation of Wasserstein GANs to train the decoder.”