AI Paper Atlas

Research map / Image generation / GANs

SONA: Learning Conditional, Unconditional, and Matching-Aware Discriminator

Yuhta Takida, Satoshi Hayakawa, Takashi Shibuya, Masaaki Imaizumi, Naoki Murata, Bac Nguyen, Toshimitsu Uesaka, Chieh-Hsin Lai, Yuki Mitsufuji

ICLR 2026 · Image generation › GANs · approach: Stabilized GAN training

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
add orthogonal projection in the alignment term (ablation)SONA · CIFAR10FID: 7.51 -> 6.29Improved
add the mismatch loss BT-mm (ablation)SONA · CIFAR10IS: 9.08 -> 9.54Improved
add both orthogonal projection and mismatch loss (ablation)SONA · CIFAR10FID: 7.51 -> 5.65Improved
add adaptive weighting of the three objectives (ablation)SONA · CIFAR10FID: 7.09 -> 5.65Improved
remove mismatching loss (ablation)SONA · 2D mixture of Gaussiansclass separationHurt
use SONA instead of PD-GAN on mixture of Gaussians with many classesconditional GAN discriminator · MoG, N >= 30number of failure casesImproved
replace projection discriminator with SONA on BigGANPD-GAN · CIFAR10FID: 4.60 -> 4.24Improved
replace projection discriminator with SONAPD-GAN · ImageNet 128 batch 256FID: 29.76 -> 13.17Improved
combine SONA with DiffAugSONA · TinyImageNetFID: 16.33 -> 7.76Improved
apply SONA to GALIP discriminator with frozen CLIP text embeddingGALIP · COCOFID: 5.30 -> 4.70Improved
apply SONA to GALIPGALIP · CUBCLIP Score: 0.3310 -> 0.3342No clear effect
compare against ReACGAN on StyleGAN2 CIFAR10 FIDSONA · CIFAR10 StyleGAN2FID: 3.39 -> 3.38No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 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