This research introduces an AI distillation framework that learns physically interpretable representations of particles and fibers from optical microscopy images.

A multimodal teacher combines image information with illumination, magnification, identity and morphology. A vision-only student learns to reconstruct that structured representation from an image alone, supporting retrieval, classification and exploratory analysis.

Authors: Simiao Sun, Kenneth Ng, Lynn Lee, Astrid Harth, Asami Odate, Aggelos Katsaggelos, Manuel Ballester Matito, Nicholas Eastaugh and Marc Walton. Submitted 1 August 2026.