Graph-CNNs for RF Imaging: Learning the Electric Field Integral Equations

2025-03-18

AI Summary

This work addresses the inverse scattering problem in radio-frequency (RF) imaging for surface reconstruction. To tackle this, we propose a physics-informed deep learning framework. First, an efficient electromagnetic forward model is constructed based on the Electric Field Integral Equation (EFIE) to generate high-fidelity synthetic training data. Second, we design a hybrid network architecture integrating Graph Attention Networks (GAT), residual CNNs, and U-Net—marking the first application of GAT to RF inverse problems to explicitly encode receiver geometric topology. Furthermore, we establish a co-optimization framework jointly training the forward model and the inverse network, balancing physical interpretability with data-driven performance. Evaluated on two heterogeneous synthetic datasets, our method achieves significantly improved reconstruction accuracy, exhibits strong robustness to measurement noise and receiver configuration variations, and maintains low inference latency—enabling real-time RF imaging.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Generative Adversarial Networks (GANs) for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Radio-Frequency (RF) imaging concerns the digital recreation of the surfaces of scene objects based on the scattered field at distributed receivers. To solve this difficult inverse scattering problems, data-driven methods are often employed that extract patterns from similar training examples, while offering minimal latency. In this paper, we first provide an approximate yet fast electromagnetic model, which is based on the electric field integral equations, for data generation, and subsequently propose a Deep Neural Network (DNN) architecture to learn the corresponding inverse model. A graph-attention backbone allows for the system geometry to be passed to the DNN, where residual convolutional layers extract features about the objects, while a UNet head performs the final image reconstruction. Our quantitative and qualitative evaluations on two synthetic data sets of different characteristics showcase the performance gains of thee proposed advanced architecture and its relative resilience to signal noise levels and various reception configurations.

Problem

• Solving inverse scattering problems in RF imaging
• Developing a fast electromagnetic model for data generation
• Proposing a DNN architecture for image reconstruction

Innovation

• Graph-attention backbone for geometry integration
• Residual convolutional layers for feature extraction
• UNet head for final image reconstruction