AI Paper Atlas

Research map / Image generation / GANs

High-Fidelity Synthesis with Disentangled Representation

Wonkwang Lee, Donggyun Kim, Seunghoon Hong, Honglak Lee

ECCV 2020 · Image generation › GANs · approach: Large-scale GAN latent analysis · 65 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
train encoder and generator in two successive steps instead of end-to-end (ablation)ID-GAN · dSpritesFactorVAE metric: 0.50 -> 0.65Improved
train encoder and generator in two successive steps instead of end-to-end (ablation)ID-GAN · dSpritesMIG: 0.13 -> 0.28Improved
train encoder and generator in two successive steps instead of end-to-end (ablation)ID-GAN · dSpritesFID: 3.18 -> 2.00Improved
use two-step ID-GAN instead of joint VAE-GAN (ablation)beta-VAE · dSpritesMIG: 0.13 -> 0.28Improved
add adversarial loss to the encoder update (VAE-GAN vs beta-VAE) (ablation)beta-VAE · dSpritesFactorVAE metric: 0.65 -> 0.46Hurt
add adversarial loss to the encoder update (VAE-GAN vs beta-VAE) (ablation)beta-VAE · dSpritesFID: 37.75 -> 33.54Improved
add distillation regularizer R_ours to generator training (ablation)ID-GAN · CelebA 128x128R_ours (alignment): -65.84 -> -10.25Improved
add distillation regularizer R_ours to generator training (ablation)ID-GAN · CelebA 128x128GILBO: -20.40 -> -0.19Improved
add distillation regularizer R_ours to generator training (ablation)ID-GAN · CelebA 128x128FID: 5.75 -> 6.61No clear effect
replace implicit cGAN alignment with explicit distillationtwo-step conditional generation · CelebA 128x128GILBO: -7.57 -> -0.19Improved
replace beta-VAE with ID-GAN on beta-VAE codebeta-VAE · CelebA (64x64)FID: 166.01 -> 4.08Improved
replace InfoGAN information maximization with distilled VAE encoderInfoGAN · 3D Chairs (64x64)FID: 60.45 -> 25.44Improved

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