FFDNet: Toward a Fast and Flexible Solution for CNN based Image Denoising

Abstract

Due to the fast inference and good performance, discriminative learning methods have been widely studied in image denoising. However, these methods mostly learn a specific model for each noise level, and require multiple models for denoising images with different noise levels. They also lack flexibility to deal with spatially variant noise, limiting their applications in practical denoising. To address these issues, we present a fast and flexible denoising convolutional neural network, namely FFDNet, with a tunable noise level map as the input. The proposed FFDNet works on downsampled sub-images, achieving a good trade-off between inference speed and denoising performance. In contrast to the existing discriminative denoisers, FFDNet enjoys several desirable properties, including (i) the ability to handle a wide range of noise levels (i.e., [0, 75]) effectively with a single network, (ii) the ability to remove spatially variant noise by specifying a non-uniform noise level map, and (iii) faster speed than benchmark BM3D even on CPU without sacrificing denoising performance. Extensive experiments on synthetic and real noisy images are conducted to evaluate FFDNet in comparison with state-of-the-art denoisers. The results show that FFDNet is effective and efficient, making it highly attractive for practical denoising applications.

IEEE Transactions on Image Processing, code: https://github.com/cszn/FFDNet

Cited by in corpus (106)

• Artificial Intelligence in the Creative Industries: A Review
• R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network
• DeepEDN: A Deep Learning-based Image Encryption and Decryption Network for Internet of Medical Things
• On the use of deep learning for phase recovery
• Practical Blind Image Denoising via Swin-Conv-UNet and Data Synthesis
• SUNet: Swin Transformer UNet for Image Denoising
• Enhanced CNN for image denoising
• WINNet: Wavelet-inspired Invertible Network for Image Denoising
• Blind Universal Bayesian Image Denoising with Gaussian Noise Level Learning
• DVDnet: A Fast Network for Deep Video Denoising
• Can learning from natural image denoising be used for seismic data interpolation?
• Path-Restore: Learning Network Path Selection for Image Restoration
• Deep Plug-and-play Prior for Low-rank Tensor Completion
• Implicit Dual-domain Convolutional Network for Robust Color Image Compression Artifact Reduction
• Interpreting and Improving Adversarial Robustness of Deep Neural Networks with Neuron Sensitivity
• Review: Deep Learning in Electron Microscopy
• Deep Learning Methods For Synthetic Aperture Radar Image Despeckling: An Overview Of Trends And Perspectives
• Large-scale phase retrieval
• COLA-Net: Collaborative Attention Network for Image Restoration
• NLH: A Blind Pixel-level Non-local Method for Real-world Image Denoising
• Noise2Fast: Fast Self-Supervised Single Image Blind Denoising
• Deep Learning Methods for Solving Linear Inverse Problems: Research Directions and Paradigms
• Joint Face Image Restoration and Frontalization for Recognition
• Deep Plug-and-Play Prior for Hyperspectral Image Restoration
• Variational Deep Image Restoration
• RE-MIMO: Recurrent and Permutation Equivariant Neural MIMO Detection
• Coupling Model-Driven and Data-Driven Methods for Remote Sensing Image Restoration and Fusion
• All-optical image denoising using a diffractive visual processor
• One-Two-One Networks for Compression Artifacts Reduction in Remote Sensing
• SIMBA: Scalable Inversion in Optical Tomography using Deep Denoising Priors
• Learning a Single Model with a Wide Range of Quality Factors for JPEG Image Artifacts Removal
• Progressive Joint Low-light Enhancement and Noise Removal for Raw Images
• A Survey on All-in-One Image Restoration: Taxonomy, Evaluation and Future Trends
• Two-Step Phase Shifting Algorithms: Where Are We?
• Color Image and Multispectral Image Denoising Using Block Diagonal Representation
• MWFormer: Multi-Weather Image Restoration Using Degradation-Aware Transformers
• Early Exit or Not: Resource-Efficient Blind Quality Enhancement for Compressed Images
• Learning Spatial and Spatio-Temporal Pixel Aggregations for Image and Video Denoising
• Unified Single-Image and Video Super-Resolution via Denoising Algorithms
• Time-Travel Rephotography
• Solar image denoising with convolutional neural networks
• End-to-end Alternating Optimization for Real-World Blind Super Resolution
• DCT2net: an interpretable shallow CNN for image denoising
• Bayesian Image Super-Resolution with Deep Modeling of Image Statistics
• LIDIA: Lightweight Learned Image Denoising with Instance Adaptation
• Deep Learning based Switching Filter for Impulsive Noise Removal in Color Images
• A Novel Image Denoising Algorithm Using Concepts of Quantum Many-Body Theory
• CDLNet: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing
• High Perceptual Quality Image Denoising with a Posterior Sampling CGAN
• Stimulating Diffusion Model for Image Denoising via Adaptive Embedding and Ensembling
• A deep convolutional neural network for salt-and-pepper noise removal using selective convolutional blocks
• Solving Inverse Problems by Joint Posterior Maximization with Autoencoding Prior
• Fast, nonlocal and neural: a lightweight high quality solution to image denoising
• Blind Inverse Gamma Correction with Maximized Differential Entropy
• 10-mega pixel snapshot compressive imaging with a hybrid coded aperture
• Selective Residual M-Net for Real Image Denoising
• Weak-signal extraction enabled by deep-neural-network denoising of diffraction data
• Deep learning-based denoising for fast time-resolved flame emission spectroscopy in high-pressure combustion environment
• Burst Photography for Learning to Enhance Extremely Dark Images
• Attention Cube Network for Image Restoration
• RDFNet: Regional Dynamic FISTA-Net for Spectral Snapshot Compressive Imaging
• Fast Diffusion EM: a diffusion model for blind inverse problems with application to deconvolution
• Combining Progressive Rethinking and Collaborative Learning: A Deep Framework for In-Loop Filtering
• Joint Multi-Scale Tone Mapping and Denoising for HDR Image Enhancement
• Securing Fixed Neural Network Steganography
• A New Recurrent Plug-and-Play Prior Based on the Multiple Self-Similarity Network
• Toward DNN of LUTs: Learning Efficient Image Restoration with Multiple Look-Up Tables
• On Maximum-a-Posteriori estimation with Plug & Play priors and stochastic gradient descent
• Efficient Burst Raw Denoising with Variance Stabilization and Multi-frequency Denoising Network
• Image Denoising using Attention-Residual Convolutional Neural Networks
• Joint Demosaicking and Denoising Benefits from a Two-stage Training Strategy
• Distributed Learning and Inference with Compressed Images
• A Multi-Head Convolutional Neural Network With Multi-path Attention improves Image Denoising
• PoGaIN: Poisson-Gaussian Image Noise Modeling from Paired Samples
• SSP-IR: Semantic and Structure Priors for Diffusion-based Realistic Image Restoration
• GIA-Net: Global Information Aware Network for Low-light Imaging
• Rank-One Network: An Effective Framework for Image Restoration
• Contextual colorization and denoising for low-light ultra high resolution sequences
• Tracking and fast imaging of a moving object via Fourier modulation
• Nonlinear Power Method for Computing Eigenvectors of Proximal Operators and Neural Networks
• Deep Camera Obscura: An Image Restoration Pipeline for Lensless Pinhole Photography
• Zero-Shot Image Denoising for High-Resolution Electron Microscopy
• Ultra-low switching current density in all-amorphous W-Hf / CoFeB / TaOx films
• Linear Combinations of Patches are Unreasonably Effective for Single-Image Denoising
• Little Strokes Fell Great Oaks: Boosting the Hierarchical Features for Multi-exposure Image Fusion
• Towards a unified view of unsupervised non-local methods for image denoising: the NL-Ridge approach
• Representing Noisy Image Without Denoising
• Enhancement of a CNN-Based Denoiser Based on Spatial and Spectral Analysis
• Robust Hyperspectral Image Fusion with Simultaneous Guide Image Denoising via Constrained Convex Optimization
• NIR-Assisted Image Denoising: A Selective Fusion Approach and A Real-World Benchmark Dataset
• Modeling Dual-Exposure Quad-Bayer Patterns for Joint Denoising and Deblurring
• Deep Learning Assisted Denoising of Experimental Micrographs
• IMAGO: A family photo album dataset for a socio-historical analysis of the twentieth century
• Evaluating Similitude and Robustness of Deep Image Denoising Models via Adversarial Attack
• Joint multi-dimensional dynamic attention and transformer for general image restoration
• Video Segmentation Learning Using Cascade Residual Convolutional Neural Network
• Pivotal Auto-Encoder via Self-Normalizing ReLU
• Removing Radio Frequency Interference from Auroral Kilometric Radiation with Stacked Autoencoders
• Towards Controllable Real Image Denoising with Camera Parameters
• Adversarial Transferability in Deep Denoising Models: Theoretical Insights and Robustness Enhancement via Out-of-Distribution Typical Set Sampling
• ReTiDe: Real-Time Denoising for Energy-Efficient Motion Picture Processing with FPGAs
• Joint multiband deconvolution for Euclid and Vera C. Rubin images
• A Research and Strategy of Remote Sensing Image Denoising Algorithms
• Hybrid Training of Denoising Networks to Improve the Texture Acutance of Digital Cameras
• Self-Supervised Noise Adaptive MRI Denoising via Repetition to Repetition (Rep2Rep) Learning
• Deep Unfolding-Aided Parameter Tuning for Plug-and-Play-Based Video Snapshot Compressive Imaging