Unrolled Optimization with Deep Priors

TL;DR

This paper introduces a framework combining classical iterative optimization with deep learning priors to improve solutions for various inverse imaging problems, outperforming existing methods.

Contribution

It presents unrolled optimization with deep priors, integrating physical models into deep networks for inverse imaging tasks, demonstrating significant performance gains.

Findings

Outperforms state-of-the-art methods in denoising, deblurring, and MRI reconstruction.

Provides insights into how and why the framework achieves superior results.

Applicable across a wide range of inverse imaging problems.

Abstract

A broad class of problems at the core of computational imaging, sensing, and low-level computer vision reduces to the inverse problem of extracting latent images that follow a prior distribution, from measurements taken under a known physical image formation model. Traditionally, hand-crafted priors along with iterative optimization methods have been used to solve such problems. In this paper we present unrolled optimization with deep priors, a principled framework for infusing knowledge of the image formation into deep networks that solve inverse problems in imaging, inspired by classical iterative methods. We show that instances of the framework outperform the state-of-the-art by a substantial margin for a wide variety of imaging problems, such as denoising, deblurring, and compressed sensing magnetic resonance imaging (MRI). Moreover, we conduct experiments that explain how the framework is best used and why it outperforms previous methods.