Artificial Intelligence for the Characterization of Particles and Fibers by Optical Microscopy

2026-07-31

AI Summary

This study addresses the challenge of semantic parsing in optical microscopy images of particle–fiber dispersion systems, where multiple imaging factors impede reliable information extraction. The authors propose a semantic-anchor-based knowledge distillation framework: a multimodal teacher model integrates image data with textual metadata—such as illumination conditions, magnification, and sample identity—to generate structured semantic vectors, while a student Vision Transformer (ViT) learns to reconstruct these vectors from images alone, thereby achieving visual–semantic alignment. The approach innovatively incorporates an interpretable block-structured teacher vector and a clustering-based pseudo-class regularizer that operates without negative sampling, effectively mitigating embedding collapse. Experimental results demonstrate that the method achieves 80% pseudo-class accuracy and 75% Recall@1 in leave-one-out nearest-neighbor retrieval, significantly enhancing fine-grained sample retrieval and classification performance.

Abstract

Optical microscopy of particle and fiber dispersions involves interpreting subtle visual cues influenced by specimen morphology, chemical composition, magnification, and illumination conditions. We introduce an artificial intelligence (AI) distillation framework that extracts semantically rich image embeddings from microscopy images using semantic anchors. A multimodal teacher combines each image's visual embedding with three text embeddings representing illumination modality, magnification, and specimen identity and morphology. Generated by LongCLIP's extended-context text encoder, this yields a 2304-dimensional block-structured teacher vector whose component blocks remain physically interpretable throughout training and inference. A student vision transformer (ViT) with a multi-layer perceptron (MLP) decoder is trained to reconstruct this teacher vector from the image alone, minimizing a mean absolute error (L1) loss that enforces coordinate-level fidelity to the teacher's block structure. A cross-entropy term over pseudo-classes derived from HDBSCAN clustering of the teacher embedding space acts as a collapse-prevention regularizer, enforcing inter-cluster separation without requiring contrastive negative mining. At inference, the student operates on image input alone, producing compact embeddings that recover the full semantic content of the teacher vector. The framework achieves approximately 80% pseudo-class validation accuracy and 75% Recall@1 on fine-grained specimen description labels under leave-one-out nearest-neighbor retrieval. These results demonstrate that semantic anchoring enables a vision-only student to acquire richer and more interpretable representations than image-only training, with direct applicability to retrieval, classification, and exploratory analysis of heterogeneous particle and fiber dispersions.

Problem

• particle characterization
• fiber analysis
• optical microscopy
• semantic representation
• image interpretation

Innovation

• semantic anchoring
• AI distillation
• multimodal teacher-student framework
• interpretable embeddings
• HDBSCAN regularization