ControlLight: Towards Controllable, Consistent, and Generalizable Low-Light Enhancement
Summary
ControlLight is a controllable low-light enhancement framework that uses a large-scale real-world dataset and a weighted flow matching loss to achieve consistent image quality across varying enhancement strengths, achieving state-of-the-art performance.
View Cached Full Text
Cached at: 05/26/26, 06:42 AM
Paper page - ControlLight: Towards Controllable, Consistent, and Generalizable Low-Light Enhancement
Source: https://huggingface.co/papers/2605.25569
Abstract
ControlLight is a controllable low-light enhancement framework that uses a large-scale real-world dataset and weighted flow matching loss to ensure consistent image quality across varying enhancement strengths.
Existing deep learning-basedlow-light enhancementmethods are typically trained on limited datasets with single enhancement targets, which restricts theirgeneralizationability andcontrollabilityin real-world applications. To overcome these limitations, we propose ControlLight, a controllable, consistent, and generalizable framework forlow-light enhancement. We first construct a large-scale dataset ofreal-world degraded imageswithcontinuous illumination-strength supervision. To further ensure consistent outputs under different control strengths, we introduce a misalignment-awareweighted flow matching lossthat preserves image structure across continuous enhancement strengths. ControlLight allows users to edit real-world degraded low-light images toward satisfactory enhancement results by flexibly controlling the strength while preserving visual consistency and realism. Extensive experiments show that ControlLight achieves state-of-the-art performance against existinglow-light enhancementapproaches while demonstrating strong continuouscontrollabilityandgeneralizationto real-world scenarios.
View arXiv pageView PDFProject pageGitHub3Add to collection
Community
Paper submitter
This comment has been hidden (marked as Graphic Content)
Paper submitter
Paper submitter
Upload images, audio, and videos by dragging in the text input, pasting, orclicking here.
Tap or paste here to upload images
Get this paper in your agent:
hf papers read 2605\.25569
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper1
#### ControlLight/ControlLight Updatedabout 2 hours ago
Datasets citing this paper1
#### ControlLight/Light100K Viewer• Updatedabout 2 hours ago • 17.8k • 36
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2605.25569 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
WildRelight: A Real-World Benchmark and Physics-Guided Adaptation for Single-Image Relighting
This paper introduces WildRelight, a new real-world benchmark dataset for single-image relighting that addresses the gap between synthetic and natural scenes. It proposes a physics-guided adaptation framework using diffusion posterior sampling and test-time adaptation to improve model performance on real-world data.
M2Retinexformer: Multi-Modal Retinexformer for Low-Light Image Enhancement
M2Retinexformer extends the Retinexformer framework for low-light image enhancement by incorporating depth, luminance, and semantic cues via cross-attention and adaptive gating, achieving state-of-the-art results on multiple benchmarks.
microsoft/Lens
Microsoft releases Lens, a 3.8B-parameter foundational text-to-image model designed for efficient training and fast high-resolution generation, achieving competitive quality with reduced compute.
Relit-LiVE: Relight Video by Jointly Learning Environment Video
This paper introduces Relit-LiVE, a novel video relighting framework that produces physically consistent results without requiring camera pose information by using raw reference images and joint environment video prediction.
PixCon: Clean-Positive Contrastive Learning for Foundation-Model Semi-Supervised Segmentation
PixCon proposes a clean-positive pixel-contrastive framework for semi-supervised semantic segmentation that guarantees contamination-free positive sets via per-class memory banks, improving accuracy over existing methods on benchmarks like Pascal VOC, Cityscapes, and ADE20K.