Iterative Residual Network for Deep Joint Image Demosaicking and Denoising

Modern digital cameras rely on sequential execution of separate image processing steps to produce realistic images. The first two steps are usually related to denoising and demosaicking where the former aims to reduce noise from the sensor and the latter converts a series of light intensity readings to color images. Modern approaches try to jointly solve these problems, i.e joint denoising-demosaicking which is an inherently ill-posed problem given that two-thirds of the intensity information are missing and the rest are perturbed by noise. While there are several machine learning systems that have been recently introduced to tackle this problem, in this work we propose a novel algorithm which is inspired by powerful classical image regularization methods, large-scale optimization and deep learning techniques. Consequently, our derived iterative neural network has a transparent and clear interpretation compared to other black-box data driven approaches. The extensive comparisons that we report demonstrate the superiority of our proposed network, which outperforms any previous approaches on both noisy and noise-free data across many different datasets using less training samples. This improvement in reconstruction quality is attributed to the principled way we design and train our network architecture, which as a result requires fewer trainable parameters than the current state-of-the-art solution.