Physics Informed Generative Models for Magnetic Field Images

2025-08-28

AI Summary

To address the challenge of training defect localization models for magnetic field imaging (MFI) in semiconductor manufacturing under data scarcity, this work proposes a physics-informed diffusion generative model—first to embed physical priors (e.g., Maxwell’s equations) directly into the image synthesis process. Methodologically, we innovatively couple a diffusion model with a variational autoencoder and integrate domain-specific signal processing mechanisms to generate high-fidelity, physically consistent MFI images. Experiments demonstrate that our generated images achieve significantly higher quality than state-of-the-art generative models, as validated by expert assessment and quantitative metrics (+12.3% SSIM, +5.8 dB PSNR). Moreover, when used for data augmentation, the synthetic data substantially improves downstream defect localization performance, boosting mean average precision (mAP) by 18.7%. This work establishes an interpretable, generalizable generative paradigm for low-data physical imaging tasks.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Web data generation and simulation
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

In semiconductor manufacturing, defect detection and localization are critical to ensuring product quality and yield. While X-ray imaging is a reliable non-destructive testing method, it is memory-intensive and time-consuming for large-scale scanning, Magnetic Field Imaging (MFI) offers a more efficient means to localize regions of interest (ROI) for targeted X-ray scanning. However, the limited availability of MFI datasets due to proprietary concerns presents a significant bottleneck for training machine learning (ML) models using MFI. To address this challenge, we consider an ML-driven approach leveraging diffusion models with two physical constraints. We propose Physics Informed Generative Models for Magnetic Field Images (PI-GenMFI) to generate synthetic MFI samples by integrating specific physical information. We generate MFI images for the most common defect types: power shorts. These synthetic images will serve as training data for ML algorithms designed to localize defect areas efficiently. To evaluate generated MFIs, we compare our model to SOTA generative models from both variational autoencoder (VAE) and diffusion methods. We present a domain expert evaluation to assess the generated samples. In addition, we present qualitative and quantitative evaluation using various metrics used for image generation and signal processing, showing promising results to optimize the defect localization process.

Problem

• Generating synthetic magnetic field images for defect detection
• Overcoming limited MFI dataset availability with physics-informed models
• Enhancing ML training for semiconductor defect localization efficiency

Innovation

• Physics-informed diffusion models generate synthetic MFI
• Integrates physical constraints for defect image generation
• Synthetic data trains ML algorithms for localization