Joint Demosaicing and Denoising with Double Deep Image Priors

TL;DR

This paper introduces JDD-DoubleDIP, a novel method for joint demosaicing and denoising of RAW images that operates without training data, outperforming existing methods on standard datasets.

Contribution

The paper presents a training-free joint demosaicing and denoising approach using Double Deep Image Priors, addressing limitations of data-dependent neural networks.

Findings

Outperforms existing methods in PSNR and SSIM

Effective on various noise types and intensities

Operates directly on single RAW images without training

Abstract

Demosaicing and denoising of RAW images are crucial steps in the processing pipeline of modern digital cameras. As only a third of the color information required to produce a digital image is captured by the camera sensor, the process of demosaicing is inherently ill-posed. The presence of noise further exacerbates this problem. Performing these two steps sequentially may distort the content of the captured RAW images and accumulate errors from one step to another. Recent deep neural-network-based approaches have shown the effectiveness of joint demosaicing and denoising to mitigate such challenges. However, these methods typically require a large number of training samples and do not generalize well to different types and intensities of noise. In this paper, we propose a novel joint demosaicing and denoising method, dubbed JDD-DoubleDIP, which operates directly on a single RAW image without requiring any training data. We validate the effectiveness of our method on two popular datasets -- Kodak and McMaster -- with various noises and noise intensities. The experimental results show that our method consistently outperforms other compared methods in terms of PSNR, SSIM, and qualitative visual perception.