AI Paper Atlas

Research map / Image generation / Image restoration

LucidFlux: Caption-Free Universal Image Restoration via a Large-Scale Diffusion Transformer

Song Fei, Tian Ye, Lujia Wang, Lei Zhu

ICLR 2026 · Image generation › Image restoration · approach: Efficient restoration architectures

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 caption-free SigLIP semantic alignment (ablation)LucidFlux dual-branch conditioner · RealLQ250MUSIQ: 61.582 -> 62.000Improved
add timestep- and layer-adaptive modulation (TLCM) (ablation)LucidFlux · RealLQ250MUSIQ: 62.000 -> 65.500Improved
scale to curated large-scale high-quality data (ablation)LucidFlux · RealLQ250MUSIQ: 65.500 -> 73.0088Improved
fully fine-tune the Flux backbone instead of freezing it (ablation)LucidFlux · RealLQ250MUSIQ: 73.01 -> 63.30Hurt
update only attention layers instead of freezing backbone (ablation)LucidFlux · RealLQ250MUSIQ: 73.01 -> 66.27Hurt
use GT captions instead of caption-free SigLIP (ablation)LucidFlux (Flux.1) · RealSRMUSIQ: 70.2005 -> 70.1654No clear effect
use VLM captions instead of caption-free SigLIP (ablation)LucidFlux (Flux.1) · RealSRNIQE: 4.2893 -> 4.6249Hurt
use VLM captions instead of caption-free (ablation)LucidFlux (Flux.1) · RealSRadded latency (s): +0.012 -> +10.057sHurt
keep top 10% by CLIP-IQA instead of 20% (ablation)LucidFlux data curation · RealLQ250MUSIQ: 73.0088 -> 69.8807Hurt
relax flat-region filter to 600 and 60% (ablation)LucidFlux data curation · RealLQ250MUSIQ: 73.0088 -> 70.2861Hurt
stricter blur range 200-6000 instead of 150-8000 (ablation)LucidFlux data curation · RealLQ250MUSIQ: 73.0088 -> 71.3748Hurt
use LucidFlux instead of DiffBIRdiffusion restoration · DIV2K-ValPSNR: 20.0389 -> 15.4393Hurt

Later papers that compared with this method

1 later accepted paper reports LucidFlux as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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