AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Mitigating Noise Shift in Denoising Generative Models with Noise Awareness Guidance

Jincheng Zhong, Boyuan Jiang, Xin Tao, Pengfei Wan, Kun Gai, Mingsheng Long

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Denoising diffusion probabilistic models

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 NAG to converged DiT-XL/2 sampling without CFGDiT-XL/2 · ImageNet 256x256FID: 9.62 -> 2.59Improved
add NAG to converged SiT-XL/2 sampling without CFGSiT-XL/2 · ImageNet 256x256FID: 8.61 -> 2.26Improved
add NAG on top of CFGSiT-XL/2 · ImageNet 256x256FID: 2.06 -> 1.72Improved
add NAG on top of CFGDiT-XL/2 · ImageNet 256x256FID: 2.27 -> 2.14Improved
add NAG without CFGDiT-XL/2 · ImageNet 256x256recall: 0.67 -> 0.60Hurt
add NAG to fine-tuned DiT-XL/2 without CFGDiT-XL/2 · 7 fine-tuning datasetsaverage FID: 18.65 -> 13.72Improved
add NAG on top of CFG in fine-tuningDiT-XL/2 · 7 fine-tuning datasetsaverage FID: 13.69 -> 10.31Improved
add NAG on top of Domain Guidance (DoG)DiT-XL/2 · 7 fine-tuning datasetsaverage FID: 11.55 -> 10.05Improved
add NAG on top of DoG on Stanford CarDiT-XL/2 · Stanford CarFID: 4.38 -> 4.21No clear effect

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