Quality-factor inspired deep neural network solver for solving inverse scattering problems

2025-04-29

AI Summary

To address the strong dependence on data quality, difficulty in artifact suppression, and low reconstruction fidelity in electromagnetic inverse scattering imaging (ISI), this paper proposes a quality-factor-driven physics-informed deep solver. Methodologically, we first define a sample quality factor to enable quantitative screening and weighted sampling of training data; design a lightweight network architecture integrating residual connections and channel-wise attention, explicitly embedding electromagnetic scattering physics as prior knowledge; and formulate a composite loss function jointly enforcing data fidelity, wave-equation constraints, and structural regularization of the reconstructed solution. Experimental results on diverse numerical benchmarks demonstrate a 32% reduction in reconstruction error and a 41% improvement in background artifact suppression, achieving state-of-the-art performance in comprehensive evaluation.

Technology Category

• Search and Optimization: Learning to Search
• Constraint Satisfaction and Optimization: Satisfiability
• Computer Vision: Learning & Optimization for CV

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Deep neural networks have been applied to address electromagnetic inverse scattering problems (ISPs) and shown superior imaging performances, which can be affected by the training dataset, the network architecture and the applied loss function. Here, the quality of data samples is cared and valued by the defined quality factor. Based on the quality factor, the composition of the training dataset is optimized. The network architecture is integrated with the residual connections and channel attention mechanism to improve feature extraction. A loss function that incorporates data-fitting error, physical-information constraints and the desired feature of the solution is designed and analyzed to suppress the background artifacts and improve the reconstruction accuracy. Various numerical analysis are performed to demonstrate the superiority of the proposed quality-factor inspired deep neural network (QuaDNN) solver and the imaging performance is finally verified by experimental imaging test.

Problem

• Optimizing training dataset quality for inverse scattering problems
• Enhancing feature extraction with advanced network architecture
• Designing loss function to improve reconstruction accuracy

Innovation

• Quality factor optimizes training dataset composition
• Residual connections and channel attention enhance features
• Multi-term loss function reduces artifacts and improves accuracy