Nonlocality-Reinforced Convolutional Neural Networks for Image Denoising

TL;DR

This paper presents a modular image denoising framework combining CNN-based local denoising with nonlocal filtering, achieving state-of-the-art results without complex network modifications.

Contribution

It introduces a nonlocality-reinforced deep CNN framework that uses standard pre-trained CNNs and nonlocal filters in a modular fashion for improved denoising.

Findings

Achieves state-of-the-art denoising performance on large grayscale datasets.

Uses a modular approach combining CNNs with nonlocal filters.

Demonstrates superior results compared to complex neural network models.

Abstract

We introduce a paradigm for nonlocal sparsity reinforced deep convolutional neural network denoising. It is a combination of a local multiscale denoising by a convolutional neural network (CNN) based denoiser and a nonlocal denoising based on a nonlocal filter (NLF) exploiting the mutual similarities between groups of patches. CNN models are leveraged with noise levels that progressively decrease at every iteration of our framework, while their output is regularized by a nonlocal prior implicit within the NLF. Unlike complicated neural networks that embed the nonlocality prior within the layers of the network, our framework is modular, it uses standard pre-trained CNNs together with standard nonlocal filters. An instance of the proposed framework, called NN3D, is evaluated over large grayscale image datasets showing state-of-the-art performance.