AI Paper Atlas

Research map / Image generation / GANs

OASIS: Only Adversarial Supervision for Semantic Image Synthesis

Vadim Sushko, Edgar Schönfeld, Dan Zhang, Jüergen Gall, Bernt Schiele, Anna Khoreva

IJCV 2022 · Image generation › GANs · approach: Large-scale GAN latent analysis · 210 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 SPADE+ discriminator with the OASIS segmentation discriminator (ablation)SPADE+ · ADE20KFID and mIoU: improves FID and mIoU by more than 30 pointsImproved
use lighter OASIS generator (remove first residual block) (ablation)SPADE+ with OASIS D · ADE20KFID and mIoU: negligible degradation of 0.3 in FID and 0.5 in mIoUNo clear effect
add LabelMix consistency regularization (ablation)OASIS · CityscapesFID: 51.5 -> 47.7Improved
add LabelMix consistency regularization (ablation)OASIS · CityscapesmIoU: 66.3 -> 69.3Improved
use CutMix instead of LabelMix for consistency regularization (ablation)OASIS · CityscapesFID: 47.7 -> 52.1Hurt
add 3D noise (ablation)OASIS · ADE20KMS-SSIM (diversity): 1.0 -> 0.63Improved
add 3D noise (ablation)OASIS · COCO-StuffmIoU: 45.5 -> 44.1Hurt
add VGG perceptual loss (ablation)OASIS · CityscapesFID: 47.7 -> 46.1Improved
add VGG perceptual loss (ablation)OASIS · ADE20KFID: 28.3 -> 31.6Hurt
add VGG perceptual loss (ablation)OASIS · ADE20Kdiversity (MS-SSIM): 0.65 -> 0.88Hurt
add feature matching loss (ablation)SPADE+ (no VGG) · ADE20KFID: 60.7 -> 49.7Improved
add feature matching loss (ablation)OASIS (no VGG) · CityscapesFID: 47.7 -> 48.5Hurt

Later papers that compared with this method

9 later accepted papers report OASIS as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.

1 more not listed.

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