AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Generative Visual Prompt: Unifying Distributional Control of Pre-Trained Generative Models

Chen Henry Wu, Saman Motamed, Shaunak Srivastava, Fernando De la Torre

NeurIPS 2022 · Image generation › Fast diffusion sampling · approach: Distributional control of diffusion · 11 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
apply PromptGen moment constraint to StyleGAN2StyleGAN2 FFHQ · race de-biasingKL divergence (race): 0.860 -> 0.099Improved
increase moment-constraint scale lambda from 1 to 2 (ablation)PromptGen · FFHQ race and ageKL divergence (race): 0.286 -> 0.099Improved
increase moment-constraint scale lambda from 1 to 2 (ablation)PromptGen · MetFaces genderKL divergence (gender): 0.000 -> 0.005Hurt
use PromptGen instead of FairStyle for combined attributesStyleGAN2 FFHQ · age+genderKL divergence: 0.0257 -> 0.000558Improved
use PromptGen instead of FairStyle for single-attribute de-biasingStyleGAN2 FFHQ · genderKL divergence: 3.20e-7 -> 1.71e-5Hurt
amortize CLIP-energy sampling with INN instead of inference-time Langevin optimization (PPGM)StyleGAN2 · inference latencyseconds per sample: 21.5 -> 0.2Improved
use PromptGen instead of PPGM n=50 for CLIP energyStyleGAN2 · baby textCLIP energy: 0.7134 -> 0.7038Improved
use PromptGen for pose-guided generation instead of GAN-ControlStyleGAN2 · same-ID embedding distancedistance: 0.68 -> 0.45Improved
decompose a complex text control into simpler energy terms (ablation)PromptGen · bald black man with beardqualitative controlImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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