AI Paper Atlas

Research map / Image generation / Super-resolution

Decoupling Direction and Magnitude: Language-Steered Flow Matching for Super-Resolution in the Dark

Jiaxin Gao, Ziyu Yue, Yaohua Liu, Danchen Cui, Zhihui Zhao, Zhixun Su

NeurIPS 2026 · Image generation › Super-resolution · approach: Perceptual and real-world SR · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes a language-steered flow-matching approach for super-resolution in the dark that decouples update direction and magnitude to adapt integration to spatially varying exposure.

Problem

Image super-resolution in low-light (dark) conditions is challenged by extreme spatial heterogeneity: underexposed noisy regions need aggressive generative enhancement while well-exposed regions need conservative updates to avoid hallucinated textures.

Approach

Standard diffusion and flow-matching models use globally uniform integration trajectories; the paper instead decouples update direction and magnitude and uses language steering to adapt them spatially, so that under-exposed regions get aggressive generative updates and well-exposed regions get conservative ones.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add degradation language prompt (ablation)DeLang-SR · RELLISUR x2 (5K ablation)PSNR: 22.433 -> 22.986Improved
add visual degradation tokens (with attribute alignment) (ablation)DeLang-SR · RELLISUR x2 (5K ablation)PSNR: 22.398 -> 22.986Improved
add attribute alignment (ablation)DeLang-SR · RELLISUR x2 (5K ablation)PSNR: 22.516 -> 22.986Improved
add attribute alignment (ablation)DeLang-SR · RELLISUR x2 (5K ablation)LPIPS: 0.2456 -> 0.2482No clear effect
add descriptor conditioning in update-strength predictor (ablation)DeLang-SR · RELLISUR x2 (5K ablation)PSNR / ColorErr: PSNR 22.754 -> 22.986; ColorErr 0.1023 -> 0.1120Mixed
add degradation-control modules to S3Diff-based base model (ablation)DeLang-SR · RELLISUR x2 (5K ablation)PSNR: 21.358 -> 22.986Improved
use LQ-only degradation intent instead of generic or oracle intent (ablation)DeLang-SR · RELLISURqualityNo clear effect
compare against DOD on no-reference qualityDeLang-SR · RELLISUR x2CLIP-IQA: 0.4224 (DOD) vs 0.4188Mixed
use SD3-based one-step flow update vs lightweight SR backbonesDeLang-SR · Inference timeseconds: 0.088 (DOD) -> 0.448Hurt
use DeLang-SR vs best baseline at x4DeLang-SR · RELLISUR x4PSNR: 20.276 (Restormer) -> 21.259Improved

Experimental setup

Builds on: flow matching

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