AI Paper Atlas

Research map / Image generation / GANs

Alleviation of Gradient Exploding in GANs: Fake Can Be Real

Song Tao, Jia Wang

CVPR 2020 · Image generation › GANs · approach: Stabilized GAN training · 33 citations (OpenAlex)

Check 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
Apply fake-as-real selection rather than real-sample zero-centered gradient penalty aloneConventional NSGAN · CIFAR-10 (500k)Inception score: 6.94\pm 0.03 -> 7.63\pm 0.05Improved
Apply fake-as-real selection rather than real-sample zero-centered gradient penalty aloneConventional NSGAN · CIFAR-10 (500k)FID: 16.66\pm 0.14 -> 12.80\pm 0.31Improved
Apply fake-as-real selection rather than real-sample zero-centered gradient penalty aloneConventional NSGAN · CIFAR-100 (500k)Inception score: 6.92\pm 0.08 -> 7.84\pm 0.04Improved
Apply fake-as-real selection rather than real-sample zero-centered gradient penalty aloneConventional NSGAN · CIFAR-100 (500k)FID: 22.28\pm 0.45 -> 17.69\pm 0.24Improved
Apply fake-as-real selection rather than real-sample zero-centered gradient penalty aloneResNet NSGAN · ImageNet (600k)Inception score: 10.66\pm 0.11 -> 11.44\pm 0.05Improved
Apply fake-as-real selection rather than real-sample zero-centered gradient penalty aloneResNet NSGAN · ImageNet (600k)FID: 44.57\pm 0.34 -> 39.69\pm 0.57Improved
Select fake samples randomly as real ones (f=1), instead of selecting low-discriminator-output samples (ablation)FARGAN · CIFAR-10Generation performanceNo clear effect
Select low-discriminator-output fake samples from a larger candidate set (f=8) rather than random fake samples (f=1) (ablation)FARGAN · CIFAR-10Generation performanceImproved
Increase the candidate-pool multiplier f up to a sufficiently large value (e.g. f=8) (ablation)FARGAN · CIFAR-10Generation performanceImproved
Increase the number of fake-as-real samples N0 appropriately (ablation)FARGAN · CIFAR-10Generation performanceImproved
Set the number of fake-as-real samples to N0=48 (ablation)FARGAN · CIFAR-10Generation performanceHurt

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