AI Paper Atlas

Research map / Image generation / GANs

WarpedGANSpace: Finding Non-Linear RBF Paths in GAN Latent Space

Christos Tzelepis, Georgios Tzimiropoulos, Ioannis Patras

ICCV 2021 · Image generation › GANs · approach: Large-scale GAN latent analysis · 50 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 non-linear RBF warping paths instead of linear directionslatent path discovery · SN-GAN MNISTreconstructor accuracy (%): 88.0 -> 98.4Improved
use non-linear RBF warping paths instead of linear directionslatent path discovery · SN-GAN AnimeFacesreconstructor accuracy (%): 99.0 -> 99.8Improved
use non-linear RBF warping paths instead of linear directionslatent path discovery · BigGAN ImageNetreconstructor accuracy (%): 85.0 -> 92.6Improved
use non-linear RBF warping paths instead of linear directionslatent path discovery · ProgGAN CelebA-HQreconstructor accuracy (%): 90.0 -> 99.3Improved
use non-linear paths instead of linear (Voynov) directionslatent path discovery · ProgGAN yaw correlationL1-normalized correlation: 0.24 -> 0.32Improved
use non-linear paths instead of linear (Voynov) directionslatent path discovery · ProgGAN pitch correlationL1-normalized correlation: 0.25 -> 0.38Improved
use non-linear paths instead of linear (Voynov) directionslatent path discovery · ProgGAN smile correlationL1-normalized correlation: 0.57 -> 0.61Improved
use non-linear paths instead of linear (Voynov) directionslatent path discovery · ProgGAN race correlationL1-normalized correlation: 0.31 -> 0.29Hurt
use non-linear paths instead of linear (Voynov) directionslatent path discovery · ProgGAN hair correlationL1-normalized correlation: 0.36 -> 0.49Improved
use non-linear paths instead of linear (Voynov) directionslatent path discovery · ProgGAN yaw rangerange r (deg): 18.93 -> 43.66Improved
use non-linear paths instead of linear (Voynov) directionslatent path discovery · ProgGAN pitch rangerange r (deg): 8.27 -> 22.53Improved
use non-linear paths instead of GANSpace PCA directionslatent path discovery · ProgGAN yaw correlationL1-normalized correlation: 0.27 -> 0.32Improved

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