AI Paper Atlas

Research map / Image generation / GANs

Robust training of implicit generative models for multivariate and heavy-tailed distributions with an invariant statistical loss

José Manuel de Frutos, Manuel A. Vázquez, Pablo M. Olmos, Joaquín Míguez

JMLR 2026 · Image generation › GANs · approach: Generative adversarial network theory · 2 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 Gaussian latent noise with generalized Pareto noise (Pareto-ISL) (ablation)ISL · KeystrokesA_CCDF: 2.7 -> 1.4Improved
replace Gaussian latent noise with generalized Pareto noise (Pareto-ISL) (ablation)ISL · Wiki TrafficA_CCDF: 8.6 -> 1.19Improved
add Pareto-ISL instead of Pareto GANHeavy-tailed generator · KeystrokesKS: 0.013 -> 0.006Improved
replace adversarial training with Pareto-ISLGAN/WGAN/MMD-GAN · Cauchy(1,2)KSD: 0.03 -> 1.90e-3Improved
pretrain with ISL-slicing before GAN trainingGAN · 2D Circle of GaussiansKL divergence: 2.24 -> 0.46Improved
pretrain WGAN with ISLWGAN · 2D-Ring, batch size 64%HQ: 30.01 -> 99.31Improved
pretrain DCGAN with ISLDCGAN · MNISTrecall (%): 75.43 -> 96.68Improved
pretrain DCGAN with ISLDCGAN · CelebArecall: 0.834 -> 0.954Improved
pretrain DCGAN with ISLDCGAN · CelebAFID: 30.93 -> 31.64No clear effect
increase K progressively instead of fixed K=10 (ablation)ISL · N(4,2)training time (s): 9016 -> 4149Improved
increase K progressively instead of fixed K=10 (ablation)ISL · Pareto(1,1)KSD: 0.24238 -> 0.08344Improved
use ISL-slicing instead of marginal ISL (ablation)ISL · Synthetic D=100 GAN dataexecution time: 10x and 5x less execution timeImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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