Hybrid Machine Learning for Scanning Near-field Optical Spectroscopy

TL;DR

This paper introduces a hybrid neural network approach combining physics-based models and experimental data to accurately predict tip-sample interactions in near-field optical microscopy, enabling better material property extraction.

Contribution

It presents a novel hybrid neural network methodology that integrates physics models with real data for improved prediction in scanning probe microscopy.

Findings

Hybrid neural network accurately predicts tip-sample interactions.

Method improves extraction of material properties from raw data.

Approach can be extended to other microscopy techniques.

Abstract

The underlying physics behind an experimental observation often lacks a simple analytical description. This is especially the case for scanning probe microscopy techniques, where the interaction between the probe and the sample is nontrivial. Realistic modeling to include the details of the probe is always exponentially more difficult than its "spherical cow" counterparts. On the other hand, a well-trained artificial neural network based on real data can grasp the hidden correlation between the signal and sample properties. In this work, we show that, via a combination of model calculation and experimental data acquisition, a physics-infused hybrid neural network can predict the tip-sample interaction in the widely used scattering-type scanning near-field optical microscope. This hybrid network provides a long-sought solution for accurate extraction of material properties from tip-specific raw data. The methodology can be extended to other scanning probe microscopy techniques as well as other data-oriented physical problems in general.