AI Paper Atlas

Research map / Image generation / GANs

Wasserstein Divergence for GANs

Jiqing Wu, Zhiwu Huang, Janine Thoma, Dinesh Acharya, Luc Van Gool

ECCV 2018 · Image generation › GANs · approach: Stabilized GAN training · 185 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace WGAN-GP gradient penalty with Wasserstein divergence objectiveWGAN-GP · LSUN (standard training)FID: 20.3 -> 15.9Improved
use Wasserstein divergence objective instead of CTGANCTGAN · CIFAR-10 (standard training)FID: 18.6 -> 18.1Improved
use Wasserstein divergence objective instead of WGAN-GPWGAN-GP · CelebA 64x64 (ResNet)FID: 18.4 -> 15.2Improved
use Wasserstein divergence instead of WGAN-GP when batch norm is removed from ResNetWGAN-GP · CelebA 64x64 (ResNet without BN)FID: 20.3 -> 18.6Improved
use Wasserstein divergence instead of WGAN-GP with ConvNetWGAN-GP · CelebA 64x64 (ConvNet)FID: 21.2 -> 17.5Improved
use Wasserstein divergence instead of WGAN-GP with ConvNet without BNWGAN-GP · CelebA 64x64 (ConvNet without BN)FID: 24.6 -> 21.5Improved
replace the PGGAN objective with the W-div objective (PGGAN-div) at 64x64PGGAN · CelebA 64x64 (progressive)FID: 16.3 -> 16.0Improved
replace the PGGAN objective with the W-div objective at 128x128PGGAN · LSUN 128x128 (progressive)FID: 15.4 -> 15.5No clear effect
replace the PGGAN objective with the W-div objective at 256x256PGGAN · LSUN 256x256 (progressive)FID: 15.1 -> 14.9Improved
increase power p from 2 to the optimum p=6 in the divergence gradient term (ablation)WGAN-div · CelebA 64x64FIDImproved
vary coefficient k with p fixed at 6 (ablation)WGAN-div · CelebA 64x64FID: FID fluctuating mildly around 16No clear effect
change sampling strategy of P_u from real-fake interpolation (1) to sampling real points only (6) (ablation)WGAN-div · CelebA 64x64FID: 15.2 -> 14.9Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy