Non-Local Recurrent Network for Image Restoration

Abstract

Many classic methods have shown non-local self-similarity in natural images to be an effective prior for image restoration. However, it remains unclear and challenging to make use of this intrinsic property via deep networks. In this paper, we propose a non-local recurrent network (NLRN) as the first attempt to incorporate non-local operations into a recurrent neural network (RNN) for image restoration. The main contributions of this work are: (1) Unlike existing methods that measure self-similarity in an isolated manner, the proposed non-local module can be flexibly integrated into existing deep networks for end-to-end training to capture deep feature correlation between each location and its neighborhood. (2) We fully employ the RNN structure for its parameter efficiency and allow deep feature correlation to be propagated along adjacent recurrent states. This new design boosts robustness against inaccurate correlation estimation due to severely degraded images. (3) We show that it is essential to maintain a confined neighborhood for computing deep feature correlation given degraded images. This is in contrast to existing practice that deploys the whole image. Extensive experiments on both image denoising and super-resolution tasks are conducted. Thanks to the recurrent non-local operations and correlation propagation, the proposed NLRN achieves superior results to state-of-the-art methods with much fewer parameters.

arXiv commentNIPS 2018

References in corpus (10)

• Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
• Convolutional Sequence to Sequence Learning
• Trainable Nonlinear Reaction Diffusion: A Flexible Framework for Fast and Effective Image Restoration
• Image Restoration Using Very Deep Convolutional Encoder-Decoder Networks with Symmetric Skip Connections
• A simple neural network module for relational reasoning
• MemNet: A Persistent Memory Network for Image Restoration
• Non-local Neural Networks
• Learning Deep CNN Denoiser Prior for Image Restoration
• Deep Networks for Image Super-Resolution with Sparse Prior
• Image Super-Resolution via Dual-State Recurrent Networks

Cited by in corpus (31)

• Neural Image Compression via Non-Local Attention Optimization and Improved Context Modeling
• Non-local Attention Optimized Deep Image Compression
• LocalViT: Analyzing Locality in Vision Transformers
• Practical Blind Image Denoising via Swin-Conv-UNet and Data Synthesis
• Noise Conscious Training of Non Local Neural Network powered by Self Attentive Spectral Normalized Markovian Patch GAN for Low Dose CT Denoising
• SwinIR: Image Restoration Using Swin Transformer
• Efficient Multi-Stage Video Denoising with Recurrent Spatio-Temporal Fusion
• Hitchhiker's Guide to Super-Resolution: Introduction and Recent Advances
• EDVR: Video Restoration with Enhanced Deformable Convolutional Networks
• s-LWSR: Super Lightweight Super-Resolution Network
• Image Super-Resolution with Cross-Scale Non-Local Attention and Exhaustive Self-Exemplars Mining
• Salient Object Detection in Video using Deep Non-Local Neural Networks
• Fast, nonlocal and neural: a lightweight high quality solution to image denoising
• Neural Nearest Neighbors Networks
• Relational recurrent neural networks
• Enhanced Standard Compatible Image Compression Framework based on Auxiliary Codec Networks
• CVEGAN: A Perceptually-inspired GAN for Compressed Video Enhancement
• Connecting Image Denoising and High-Level Vision Tasks via Deep Learning
• Self-supervised Low Light Image Enhancement and Denoising
• Rank-One Network: An Effective Framework for Image Restoration
• GAN2GAN: Generative Noise Learning for Blind Denoising with Single Noisy Images
• Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects
• The Power of Complementary Regularizers: Image Recovery via Transform Learning and Low-Rank Modeling
• Fully Convolutional Pixel Adaptive Image Denoiser
• Skeleton-Based Action Recognition with Synchronous Local and Non-local Spatio-temporal Learning and Frequency Attention
• Finding Discriminative Filters for Specific Degradations in Blind Super-Resolution
• Posterior Sampling for Image Restoration using Explicit Patch Priors
• Towards the Automation of Deep Image Prior
• Region-Adaptive Deformable Network for Image Quality Assessment
• Fine-grained Image-to-Image Transformation towards Visual Recognition
• Learning Parallax Attention for Stereo Image Super-Resolution