AI Paper Atlas

Research map / Image generation / Super-resolution

Cascaded Diffusion Models for High Fidelity Image Generation

Jonathan Ho, Chitwan Saharia, William Chan, David J. Fleet, Mohammad Norouzi, Tim Salimans

JMLR 2022 · Image generation › Super-resolution · approach: Diffusion-based super-resolution · 604 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
add conditioning augmentation to the 16->64 super-resolution model (ablation)Cascaded diffusion pipeline · ImageNet 64x64 small-scale cascadingFID vs train: 6.02 -> 2.13Improved
cascade 16->64 with conditioning augmentation instead of no cascading (ablation)non-cascaded 64x64 baseline · ImageNet 64x64FID vs train: 2.35 -> 2.13Improved
cascade 16->64 without conditioning augmentation (ablation)non-cascaded 64x64 baseline · ImageNet 64x64FID vs train: 2.35 -> 6.02Hurt
add conditioning augmentation in 32->64 cascade (ablation)CDM 32->64 · ImageNet 64x64FID vs train: 1.71 -> 1.48Improved
use non-truncated instead of truncated conditioning augmentation (ablation)CDM 32->64 · ImageNet 64x64FID vs train: 1.48 vs 1.48No clear effect
increase truncation time s too high (1501 vs 501) (ablation)CDM 32->64 truncated · ImageNet 64x64FID vs train: 1.48 -> 1.54Hurt
add conditioning augmentation when conditioning on ground truth low-res data (ablation)32->64 super-resolution · ImageNet 64x64FID vs train: 0.76 -> 1.11Hurt
add Gaussian blur augmentation sigma U(0.4,0.6) to 64->256 model (ablation)64->256 super-resolution · ImageNet 256x256FID vs train: 7.26 -> 6.18Improved
use wider blur range sigma U(0.4,0.8) instead of U(0.4,0.6) (ablation)64->256 super-resolution · ImageNet 256x256FID vs train: 6.18 -> 6.90Hurt
add class conditioning to 64->256 super-resolution (ablation)64->256 super-resolution · ImageNet 256x256FID vs train: 6.18 -> 5.75Improved
increase training batch size from 256 to 1024 (ablation)64->256 super-resolution · ImageNet 256x256FID vs train: 5.75 -> 5.00Improved
add random left-right flips to super-resolution training (ablation)64->256 super-resolution · ImageNet 256x256FID vs train: 5.00 -> 4.88Improved

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