AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

RealUHR: Harnessing Patch-Cascade Flows for Photorealistic Ultra-High-Resolution Synthesis

Yongsheng Yu, Haitian Zheng, Zhe Lin, Connelly Barnes, Yuqian Zhou, Zhifei Zhang, Jiebo Luo

AAAI 2026 · Image generation › Autoregressive and latent generation · approach: Generative latent space tooling · 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
add non-uniform time schedule vs uniform (ablation)RealUHR · UHR benchmark 2KPatch FID: 29.72 -> 26.09Improved
add non-uniform time schedule vs cosine (ablation)RealUHR · UHR benchmark 2KQualiCLIP: 0.492 -> 0.495No clear effect
add Patch Initialization from anchor patches (ablation)RealUHR · UHR benchmark 4KFID: 145.82 -> 139.29Improved
add Patch Initialization from anchor patches (ablation)RealUHR · UHR benchmark 4KPatch FID: 53.11 -> 46.15Improved
add Guidance-Consistent Adaptation (ablation)RealUHR · UHR benchmark 4KPatch FID: 50.48 -> 46.15Improved
add GCA Stage 1 reverse distillation (ablation)RealUHR · UHR benchmark 4KFID: 141.03 -> 139.29Improved
add GCA Stage 1 reverse distillation (ablation)RealUHR · UHR benchmark 2KPatch FID: 29.03 -> 26.10Improved
add Guidance-Consistent Adaptation (ablation)RealUHR · UHR benchmark 2KFID: 42.14 -> 40.29Improved
patch-cascade pipeline vs I-Max on same FLUX backboneRealUHR · UHR benchmark 4KTime (s): 751 -> 108Improved
RealUHR vs UltraPixel on CLIPScoreRealUHR · UHR benchmark 4KCLIPScore: 34.83 -> 34.19Hurt
RealUHR vs Diffusion-4K on 2K timeRealUHR · UHR benchmark 2KTime (s): 22 -> 38Hurt

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