Deep network series for large-scale high-dynamic range imaging

TL;DR

This paper introduces a residual DNN series method for large-scale high-dynamic range imaging that efficiently reconstructs images by iteratively adding residuals, achieving high quality with fewer iterations.

Contribution

It presents a novel residual DNN series approach, interpreted as learned matching pursuit, for scalable high-dynamic range imaging that outperforms traditional iterative methods in efficiency.

Findings

Competitive reconstruction quality with fewer terms

Reduced computational cost compared to PnP methods

Effective in radio-astronomical imaging simulations

Abstract

We propose a new approach for large-scale high-dynamic range computational imaging. Deep Neural Networks (DNNs) trained end-to-end can solve linear inverse imaging problems almost instantaneously. While unfolded architectures provide robustness to measurement setting variations, embedding large-scale measurement operators in DNN architectures is impractical. Alternative Plug-and-Play (PnP) approaches, where the denoising DNNs are blind to the measurement setting, have proven effective to address scalability and high-dynamic range challenges, but rely on highly iterative algorithms. We propose a residual DNN series approach, also interpretable as a learned version of matching pursuit, where the reconstructed image is a sum of residual images progressively increasing the dynamic range, and estimated iteratively by DNNs taking the back-projected data residual of the previous iteration as input. We demonstrate on radio-astronomical imaging simulations that a series of only few terms provides a reconstruction quality competitive with PnP, at a fraction of the cost.