CARD: Correlation Aware Restoration with Diffusion

TL;DR

CARD is a novel diffusion-based image restoration method that explicitly handles correlated noise by whitening observations, improving performance on real-world sensor noise and filling a gap with a new dataset.

Contribution

We introduce CARD, a training-free extension of DDRM that effectively manages correlated Gaussian noise, and present CIN-D, a new dataset for real-world sensor noise evaluation.

Findings

CARD outperforms existing methods on synthetic correlated noise tasks.

CARD demonstrates superior results on real sensor noise in CIN-D.

The CIN-D dataset enables realistic evaluation of restoration methods.

Abstract

Denoising diffusion models have achieved state-of-the-art performance in image restoration by modeling the process as sequential denoising steps. However, most approaches assume independent and identically distributed (i.i.d.) Gaussian noise, while real-world sensors often exhibit spatially correlated noise due to readout mechanisms, limiting their practical effectiveness. We introduce Correlation Aware Restoration with Diffusion (CARD), a training-free extension of DDRM that explicitly handles correlated Gaussian noise. CARD first whitens the noisy observation, which converts the noise into an i.i.d. form. Then, the diffusion restoration steps are replaced with noise-whitened updates, which inherits DDRM's closed-form sampling efficiency while now being able to handle correlated noise. To emphasize the importance of addressing correlated noise, we contribute CIN-D, a novel correlated noise dataset captured across diverse illumination conditions to evaluate restoration methods on real rolling-shutter sensor noise. This dataset fills a critical gap in the literature for experimental evaluation with real-world correlated noise. Experiments on standard benchmarks with synthetic correlated noise and on CIN-D demonstrate that CARD consistently outperforms existing methods across denoising, deblurring, and super-resolution tasks.