AI Paper Atlas

Research map / Image generation / GANs

DRB-GAN: A Dynamic ResBlock Generative Adversarial Network for Artistic Style Transfer

Wenju Xu, Chengjiang Long, Ruisheng Wang, Guanghui Wang

ICCV 2021 · Image generation › GANs · approach: Large-scale GAN latent analysis · 127 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
use SW-LIN decoder instead of instance normalization (ablation)DRB-GAN · WikiArt ablationdeception score: 0.561 -> 0.573Improved
use SW-LIN decoder instead of layer normalization (ablation)DRB-GAN · WikiArt ablationdeception score: 0.568 -> 0.573Improved
use Dynamic ResBlock instead of AdaIN ResBlock (ablation)DRB-GAN · WikiArt ablationdeception score: 0.552 -> 0.573Improved
add VGG encoder to the style code (ablation)DRB-GAN · WikiArt ablationcontent score: 66.3% -> 72.2%Improved
add style classification loss (ablation)DRB-GAN · WikiArt ablationdeception score: 0.571 -> 0.573No clear effect
add adversarial loss (ablation)DRB-GAN · WikiArt ablationdeception score: 0.542 -> 0.573Improved
add adversarial loss (ablation)DRB-GAN · WikiArt ablationcontent score: 32.2% -> 72.2%Improved
increase dynamic blocks from 1 to 4 (ablation)DRB-GAN · WikiArtdeception score: 0.567 -> 0.573Improved
use one dynamic block instead of four AdaIN blocks (ablation)DRB-GAN · WikiArtdeception score: 0.558 -> 0.567Improved
increase collection size K from 2 to 20DRB-GAN · Collection style transferdeception score: 0.576 -> 0.583Improved
increase collection size K from 2 to 20AdaIN · Collection style transferdeception score: 0.066 -> 0.011Hurt
use collection discriminator instead of conditional discriminator (ablation)DRB-GAN · Qualitative comparisonvisual qualityImproved

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