C2: Machine Learning for Multi-Scale and Multi-Modal Microscopy Data Analysis
Machine Learning for Multi-Scale and Multi-Modal Microscopy Data Analysis
This doctoral project aims to develop advanced machine learning methods for analyzing and integrating multimodal microscopy and tomography data across scales. The candidate will adapt and fine-tune state-of-the-art self-supervised and transfer learning models for tasks such as segmentation, registration, and correlation of 2D and 3D datasets from electron, X-ray, and optical microscopy. A key focus will be on representation learning that efficiently handles limited or partially labeled data. Physics-informed pretext tasks and synthetic data will be incorporated to enhance model generalization and reduce the gap between simulated and experimental data. The resulting models will enable robust characterization of complex hierarchical materials with minimal manual intervention. This work will provide foundational tools for automated, data-driven microscopy and accelerate the understanding of structure–property relationships in functional materials.
