Fast Physics-Driven Untrained Network for Highly Nonlinear Inverse Scattering Problems

2026-02-14

AI Summary

This work addresses the computational inefficiency of training-free neural networks in solving highly nonlinear electromagnetic inverse scattering problems, which arises from high-dimensional spatial-domain optimization. To overcome this challenge, the authors propose a real-time physics-driven Fourier spectral solver that represents induced currents via a truncated Fourier basis in the spectral domain, thereby restricting optimization to a low-frequency parameter subspace. The method integrates a contraction integral equation to mitigate nonlinearity at high contrasts, employs a contrast-compensating operator to correct spectral decay, and introduces a bridging suppression loss to enhance boundary sharpness of reconstructed scatterers. Requiring no training, the proposed approach achieves approximately 100-fold acceleration over existing training-free methods while maintaining robustness to noise and antenna positioning errors, enabling sub-second, high-quality microwave imaging reconstruction.

Technology Category

• Search and Optimization: Non-convex Optimization
• Machine Learning: Optimization
• Computer Vision: Learning & Optimization for CV

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Untrained neural networks (UNNs) offer high-fidelity electromagnetic inverse scattering reconstruction but are computationally limited by high-dimensional spatial-domain optimization. We propose a Real-Time Physics-Driven Fourier-Spectral (PDF) solver that achieves sub-second reconstruction through spectral-domain dimensionality reduction. By expanding induced currents using a truncated Fourier basis, the optimization is confined to a compact low-frequency parameter space supported by scattering measurements. The solver integrates a contraction integral equation (CIE) to mitigate high-contrast nonlinearity and a contrast-compensated operator (CCO) to correct spectral-induced attenuation. Furthermore, a bridge-suppressing loss is formulated to enhance boundary sharpness between adjacent scatterers. Numerical and experimental results demonstrate a 100-fold speedup over state-of-the-art UNNs with robust performance under noise and antenna uncertainties, enabling real-time microwave imaging applications.

Problem

• inverse scattering
• untrained neural networks
• real-time imaging
• high nonlinearity
• computational efficiency

Innovation

• Untrained Neural Networks
• Inverse Scattering
• Fourier-Spectral Solver
• Contraction Integral Equation
• Real-Time Microwave Imaging