Neural Inverse Scattering with Score-based Regularization

TL;DR

This paper introduces a neural field method with score-based regularization for inverse scattering, improving image reconstruction quality by leveraging denoising score functions as priors.

Contribution

The paper presents a novel neural field approach that integrates score-based regularization for joint estimation in inverse scattering problems.

Findings

Outperforms total variation regularization in simulated high-contrast objects

Provides better imaging quality than existing neural field methods

Demonstrates flexibility in joint estimation of image and scattering field

Abstract

Inverse scattering is a fundamental challenge in many imaging applications, ranging from microscopy to remote sensing. Solving this problem often requires jointly estimating two unknowns -- the image and the scattering field inside the object -- necessitating effective image prior to regularize the inference. In this paper, we propose a regularized neural field (NF) approach which integrates the denoising score function used in score-based generative models. The neural field formulation offers convenient flexibility to performing joint estimation, while the denoising score function imposes the rich structural prior of images. Our results on three high-contrast simulated objects show that the proposed approach yields a better imaging quality compared to the state-of-the-art NF approach, where regularization is based on total variation.