Image Restoration

Follow this topic Ask about this topic

Image restoration estimates cleaner or more complete images from degraded measurements, supporting applications from photography and remote sensing to medical and scientific imaging. Methods range from inverse-problem and physics-based models to CNNs, Transformers, state-space networks, and diffusion or flow models; self-supervised learning and pretrained priors help when clean paired examples are scarce. Work increasingly targets unified models for mixed degradations, real-world adaptation, and efficient deployment, with evaluation extending beyond image metrics to downstream tasks such as detection and diagnosis. A central challenge is balancing fidelity to observed evidence with perceptual quality: generative methods can add plausible but incorrect detail, so reliable control and hallucination assessment remain important.

Latest papers

372 papers42 papers in the last four weeks, up 27% on the four weeks before. 0.2% of all new papers.

Simpler and Stronger Baselines for All-in-One Image Restoration
Dec 11, 2025Hongyu Du, Yuqi Wang, Wenlong Jiao +2All-in-One Image Restoration Image Restoration
FMA-Net++: Motion- and Exposure-Aware Joint Video Super-Resolution and Deblurring
Dec 4, 2025Geunhyuk Youk, Jihyong Oh, Munchurl KimImage Restoration Video Super-Resolution
KAIST, Republic of Korea · CMLab, Chung-Ang University, Republic of Korea
SurgiATM: A Physics-Guided Plug-and-Play Model for Deep Learning-Based Smoke Removal in Laparoscopic Surgery
Nov 7, 2025Mingyu Sheng, Jianan Fan, Dongnan Liu +3Physics-Guided Image Restoration Image Restoration
School of Computer Science, The University of Sydney, Sydney, NSW, Australia · Institute of Medical Robotics, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China · Department of Radiology, Brigham and Women’s Hospital, and Harvard Medical School, Boston, MA, United States
CharDiff-LP: A Diffusion Model with Character-Level Guidance for License Plate Image Restoration
Oct 20, 2025Kihyun Na, Gyuhwan Park, Injung KimDiffusion Models for Image Restoration Diffusion Model Guidance
Department of Computer Science and Electrical Engineering, General Graduate School, Handong Global University, Pohang 37554, Republic of Korea
$\mathbf{R}^3$ : Reconstruction, Raw, and Rain: Deraining Directly in the Bayer Domain
Sep 28, 2025Nate Rothschild, Moshe Kimhi, Avi Mendelson +1Image Deraining Image Restoration
Department of ECE, Technion, Israel · Department of CS, Technion, Israel · School of ECE, Ben-Gurion University
RestoRect: Degraded Image Restoration via Latent Rectified Flow & Feature Distillation
Sep 27, 2025Shourya Verma, Mengbo Wang, Nadia Atallah Lanman +1Image Restoration Feature-Based Knowledge Distillation
Department of Computer Science, Purdue University, USA · Purdue Institute of Cancer Research, USA
InSituRes: A Physics-Informed Same-Grid Model for Enhanced Dynamic X-ray Micro-CT Reconstructions
Aug 25, 2025Qinyi Tian, Andrea Bisciotti, Soniya Tiwari +2Image Restoration Physics-Informed ML
Department of Civil & Environmental Engineering, Duke University, Durham, North Carolina, 27708, USA · Department of Physics and Earth Sciences, University of Ferrara, Via Saragat 1, Ferrara 44141, Italy · Department of Mechanical Engineering & Materials Science, Duke University, Durham, North Carolina, 27708, USA
Freqformer: Image-Demoiréing Transformer via Effective Frequency Decomposition
May 25, 2025Xiaoyang Liu, Bolin Qiu, Zheng Chen +5Frequency-Domain Image Processing Image Restoration
Shanghai Jiao Tong University
HonestFace: Towards Honest Face Restoration with One-Step Diffusion Model
May 24, 2025Jingkai Wang, Wu Miao, Jue Gong +5Image Restoration Diffusion Models for Image Restoration
School of Computer Science, Shanghai Jiao Tong University, Shanghai, China · vivo BlueImage Lab, vivo Mobile Communication Co., Ltd., Hangzhou, China
Fast and Accurate Image Restoration and Generation with Rank Enhanced Linear Attention
May 22, 2025Yuang AiEfficient ViTs Linear Attention
University of Chinese Academy of Sciences, Beijing, China
Circular Quasiconformal Deturbulence: Geometry-Based Restoration from Multiple Turbulent Frames
Apr 18, 2025Chu Chen, Han Zhang, Lok Ming LuiImage Restoration
Department of Mathematics, City University of Hong Kong Hong Kong Centre for Cerebro-Cardiovascular Health Engineering Hong Kong · Department of Mathematics, Chinese University of Hong Kong Hong Kong
Hyperspectral Image Restoration and Super-resolution with Physics-Aware Deep Learning for Biomedical Applications
Mar 3, 2025Yuchen Xiang, Zhaolu Liu, Monica Emili Garcia-Segura +12Hyperspectral Imaging Physics-Guided Image Restoration
Department of Metabolism, Digestion and Reproduction, Imperial College London, London, UK · Department of Mathematics, Imperial College London, London, UK · Department of Clinical Neurosciences and NIHR Biomedical Research Centre, University of Cambridge, Cambridge, UK +4

Dec 11, 2025J/K move · Enter open · S save

Simpler and Stronger Baselines for All-in-One Image Restoration

Hongyu Du, Yuqi Wang, Wenlong Jiao, Wenwen Li+1

All-in-One Image Restoration Image Restoration U-Net

Open paper AskPDF arXiv

Current methods for all-in-one image restoration increasingly rely on complex mechanisms, such as Mixture-of-Experts or prompt learning, to handle diverse degradations.In this work, we challenge this trend by showing that structural symmetry provides a strong and effective inductive bias, which is often sufficient to support high-quality restoration in many practical settings. We propose SymUNet, a minimalist baseline that enforces strict encoder-decoder alignment via simple additive fusion, which outperforms complex state-of-the-art models on standard benchmarks while significantly reducing computational costs. Furthermore, to address the negative transfer observed in highly diverse degradation scenarios (e.g., 5-task settings), we introduce SE-SymUNet, which leverages frozen semantic priors (CLIP) to disentangle conflicting restoration tasks. Extensive experiments demonstrate that our approach establishes a simpler, stronger, and more efficient foundation for multi-degradation learning.