Context-Aware Deep Learning for Defect Classification in Atomic-Resolution STEM
2026-06-08 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors developed a new AI method that not only looks at electron microscope images of materials but also considers extra information like the material's composition and how the images were taken. This helps the AI better identify defects in materials because it understands the full context, not just the image contrast. Their method worked very well on simulated data and performed almost as accurately as humans on real data. By linking image details to the physical settings, their approach makes defect identification more precise and reliable.
Electron microscopyImage contrastDefect classificationTransition-metal dichalcogenidesContext-aware learningMetadataBeam energyDetector geometrySimulated datasetPosterior entropy
Authors
Jiadong Dan, Cheng Zhang, Leyi Loh, Ivan Verzhbitskiy, Yuan Chen, Goki Eda, Michel Bosman, N. Duane Loh
Abstract
Artificial intelligence is rapidly advancing materials characterization, yet most applications in electron microscopy rely solely on image contrast, overlooking the chemical and experimental context that shapes image formation. This limitation makes defect classification inherently ambiguous, as similar contrasts can arise from different materials or imaging conditions. Here we develop a context-aware learning framework that integrates image-derived contrast with metadata describing composition, beam energy, and detector geometry. Using a systematically constructed dataset of ~55 million simulated patches spanning 576 cases across 96 doped monolayer transition-metal dichalcogenides, we show that conditioning on contextual variables transforms defect classification from an ill-posed image-only task into a well-posed, physically grounded problem. The framework achieves over 98% accuracy on simulations and near-human agreement on experimental data, with a 94% reduction in posterior entropy. By emphasizing contextual grounding over architectural complexity, this approach links experimental image contrast to the underlying chemical and imaging conditions, supporting physically grounded defect assignments and a general pathway toward multimodal AI models for autonomous materials characterization.