AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Spectral Guidance for Flexible and Efficient Control of Diffusion Models

Gabriel Moreira, Manuel Marques, Joao Costeira, Chenyan Xiong

ICML 2026 · Image generation › Fast diffusion sampling · approach: Guidance for conditional diffusion · 0 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 Spectral Guidance instead of the strongest training-free baseline TFGDDPM 36M · CIFAR-10 / Labelsaccuracy: 52.0 -> 89.4Improved
use Spectral Guidance instead of TFGDDPM 36M · CIFAR-10 / LabelsFID: 91.7 -> 70.7Improved
use Spectral Guidance instead of TFGDDPM 114M · CelebA-HQ / Gender + Ageaccuracy: 75.2 -> 91.5Improved
use Spectral Guidance instead of LGDDDPM 114M · CelebA-HQ / Gender + Agelog KID: -5.10 -> -2.98Hurt
use Spectral Guidance instead of TFGDDPM 121M · ImageNet / Labelsaccuracy: 40.9 -> 41.6Improved
use Spectral Guidance instead of TFGDDPM 121M · ImageNet / LabelsFID: 176 -> 183Hurt
use Spectral Guidance instead of TFG for hair-mask guidanceDDPM 114M · CelebA-HQ / MaskIoU: 0.78 -> 0.80Improved
use Spectral Guidance instead of TFG for CLIP text guidanceDDPM 114M · CelebA-HQ / CLIPVQAScore: 0.62 -> 0.64Improved
use Spectral Guidance instead of TFG for sampling latencyDDPM 114M · CelebA-HQ computational costlatency (ms/step): 81.2 -> 21.7Improved
use Spectral Guidance instead of TFG, including offline training costDDPM 114M · CelebA-HQ computational costeffective time for 10k images (h): 22.5 -> 16.9Improved
increase spectral rank K (8 to 128 and beyond) (ablation)Spectral Guidance · CIFAR-10 / LabelsaccuracyImproved
increase spectral rank K at fixed guidance strength (ablation)Spectral Guidance · CIFAR-10 / LabelsFIDHurt

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