Noise2Noise: Learning Image Restoration without Clean Data

Abstract

We apply basic statistical reasoning to signal reconstruction by machine learning -- learning to map corrupted observations to clean signals -- with a simple and powerful conclusion: it is possible to learn to restore images by only looking at corrupted examples, at performance at and sometimes exceeding training using clean data, without explicit image priors or likelihood models of the corruption. In practice, we show that a single model learns photographic noise removal, denoising synthetic Monte Carlo images, and reconstruction of undersampled MRI scans -- all corrupted by different processes -- based on noisy data only.

Added link to official implementation and updated MRI results to match it

Cited by in corpus (60)

• Self-Supervised Learning of Physics-Guided Reconstruction Neural Networks without Fully-Sampled Reference Data
• SAR2SAR: a semi-supervised despeckling algorithm for SAR images
• RARE: Image Reconstruction using Deep Priors Learned without Ground Truth
• Review: Deep Learning in Electron Microscopy
• Neural Dual Contouring
• Artificial Intelligence-Based Image Enhancement in PET Imaging: Noise Reduction and Resolution Enhancement
• Noise2Fast: Fast Self-Supervised Single Image Blind Denoising
• An Annotation-free Restoration Network for Cataractous Fundus Images
• Deep learning in biomedical optics
• Unsupervised Deep Learning Methods for Biological Image Reconstruction and Enhancement
• Deep Learning for Retrospective Motion Correction in MRI: A Comprehensive Review
• Wasserstein GANs for MR Imaging: from Paired to Unpaired Training
• Anatomical-Guided Attention Enhances Unsupervised PET Image Denoising Performance
• Speech Denoising Without Clean Training Data: A Noise2Noise Approach
• Coordinate-based neural representations for computational adaptive optics in widefield microscopy
• List-Mode PET Image Reconstruction Using Deep Image Prior
• Natural Image Noise Dataset
• A Review on Low-Dose Emission Tomography Post-Reconstruction Denoising with Neural Network Approaches
• LIDIA: Lightweight Learned Image Denoising with Instance Adaptation
• Bayesian Image Super-Resolution with Deep Modeling of Image Statistics
• Systematic Review on Learning-based Spectral CT
• A deep convolutional neural network for salt-and-pepper noise removal using selective convolutional blocks
• Noise2Context: Context-assisted Learning 3D Thin-layer Low Dose CT Without Clean Data
• Weak-signal extraction enabled by deep-neural-network denoising of diffraction data
• Multi-temporal speckle reduction with self-supervised deep neural networks
• Deep learning-based denoising for fast time-resolved flame emission spectroscopy in high-pressure combustion environment
• CERL: A Unified Optimization Framework for Light Enhancement with Realistic Noise
• Joint Multi-Scale Tone Mapping and Denoising for HDR Image Enhancement
• Frequency Domain Loss Function for Deep Exposure Correction of Dark Images
• A Practical Solution for SAR Despeckling With Adversarial Learning Generated Speckled-to-Speckled Images
• Riesz-Quincunx-UNet Variational Auto-Encoder for Satellite Image Denoising
• Self-Supervised Training For Low Dose CT Reconstruction
• A relic sketch extraction framework based on detail-aware hierarchical deep network
• Electro-Magnetic Side-Channel Attack Through Learned Denoising and Classification
• Finding Nano-Ötzi: Semi-Supervised Volume Visualization for Cryo-Electron Tomography
• Enhancement by Your Aesthetic: An Intelligible Unsupervised Personalized Enhancer for Low-Light Images
• Neural Network Methods for Radiation Detectors and Imaging
• ELMformer: Efficient Raw Image Restoration with a Locally Multiplicative Transformer
• Astronomical image denoising by self-supervised deep learning and restoration processes
• A three-dimensional MR-STAT protocol for high-resolution multi-parametric quantitative MRI
• OpenDenoising: an Extensible Benchmark for Building Comparative Studies of Image Denoisers
• A Dataset-free Deep learning Method for Low-Dose CT Image Reconstruction
• Robust multi-coil MRI reconstruction via self-supervised denoising
• Poisson2Sparse: Self-Supervised Poisson Denoising From a Single Image
• Automatic Tuning of Denoising Algorithms Parameters Without Ground Truth
• Joint tone mapping and denoising of thermal infrared images via multi-scale Retinex and multi-task learning
• Targeted Gradient Descent: A Novel Method for Convolutional Neural Networks Fine-tuning and Online-learning
• Direct Evolutionary Optimization of Variational Autoencoders With Binary Latents
• Unpaired Image Denoising
• Cosmic Ray Rejection with Attention Augmented Deep Learning
• Rotational Augmented Noise2Inverse for Low-dose Computed Tomography Reconstruction
• Explicit Abnormality Extraction for Unsupervised Motion Artifact Reduction in Magnetic Resonance Imaging
• Face Inverse Rendering via Hierarchical Decoupling
• Inverse 3D Microscopy Rendering for Cell Shape Inference with Active Mesh
• Cross-domain Denoising for Low-dose Multi-frame Spiral Computed Tomography
• Scale-Equivariant Imaging: Self-Supervised Learning for Image Super-Resolution and Deblurring
• Deep High-Resolution Network for Low Dose X-ray CT Denoising
• Practical Deep Raw Image Denoising on Mobile Devices
• Fast strategies for multi-temporal speckle reduction of Sentinel-1 GRD images
• Machine learning application to Fermi-LAT data: sharpening all-sky map and emphasizing variable sources