AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Balanced Training for Sparse GANs

Yite Wang, Jing Wu, Naira Hovakimyan, Ruoyu Sun

NeurIPS 2023 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 6 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
adjust discriminator density with balance ratio instead of fixed strong discriminatorSDST-Strong-RigL · BigGAN CIFAR-10, G density 10%training FLOPs: 83.97\% -> 10.39\%Improved
use ADAPT_relax over second-best baselinesparse SNGAN · CIFAR-10, G density 10%FID: 2.3 FID improvementImproved
use ADAPT_relax over second-best baselinesparse SNGAN · STL-10, G density 10%FID: 43.08 -> 35.98Improved
apply DST to both generator and discriminator (DST-bothGD) instead of generator only (ablation)SDST-Strong-SET · SNGAN CIFAR-10, G density 10%FID: 16.49 -> 20.57Hurt
apply DST to both generator and discriminator (RigL) (ablation)SDST-Strong-RigL · SNGAN CIFAR-10, G density 10%FID: 17.02 -> 31.95Hurt
ADAPT_relax instead of post-hoc magnitude pruningsparse SNGAN · CIFAR-10, G density 10%FID: 20.89 -> 14.19Improved
strong discriminator strategy instead of balance strategySTATIC · SNGAN STL-10, G density 10%FID: 48.18 -> 52.48Hurt
RigL growth instead of random SET growth (ablation)SDST-Strong · SNGAN STL-10, G density 10%FID: 67.37 -> 53.65Improved
add dynamic density adjust (DDA) to STATICSTATIC SNGAN · CIFAR-10FIDImproved
SDST instead of STATIC when discriminator density is small (10%)sparse SNGAN · CIFAR-10FIDNo clear effect
ADAPT_strict instead of ADAPT_relax (ablation)ADAPT · SNGAN STL-10, G density 10%FID: 35.98 -> 41.18; 10.19 -> 9.29Mixed
ADAPT_strict instead of SDST-Strong-RigL under d_D<=50%SNGAN STL-10, G density 10% · STL-10FID: 53.74 -> 41.18Improved

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