LIDIA: Lightweight Learned Image Denoising with Instance Adaptation

TL;DR

This paper introduces LIDIA, a lightweight, learnable image denoising architecture that combines classical concepts with neural networks, achieving near state-of-the-art results with fewer parameters and adaptable to specific images.

Contribution

It presents a novel lightweight neural network architecture for image denoising that integrates classical methods and combines supervised and unsupervised training for universal and adaptive denoising.

Findings

Achieves near state-of-the-art denoising performance with fewer parameters.

Effectively adapts to specific images to boost denoising quality.

Combines classical patch-based concepts with modern neural network design.

Abstract

Image denoising is a well studied problem with an extensive activity that has spread over several decades. Despite the many available denoising algorithms, the quest for simple, powerful and fast denoisers is still an active and vibrant topic of research. Leading classical denoising methods are typically designed to exploit the inner structure in images by modeling local overlapping patches, while operating in an unsupervised fashion. In contrast, recent newcomers to this arena are supervised and universal neural-network-based methods that bypass this modeling altogether, targeting the inference goal directly and globally, while tending to be very deep and parameter heavy. This work proposes a novel lightweight learnable architecture for image denoising, and presents a combination of supervised and unsupervised training of it, the first aiming for a universal denoiser and the second for adapting it to the incoming image. Our architecture embeds in it several of the main concepts taken from classical methods, relying on patch processing, leveraging non-local self-similarity, exploiting representation sparsity and providing a multiscale treatment. Our proposed universal denoiser achieves near state-of-the-art results, while using a small fraction of the typical number of parameters. In addition, we introduce and demonstrate two highly effective ways for further boosting the denoising performance, by adapting this universal network to the input image.