Deep-Learning-Aided Extraction of Optical Constants in Scanning Near-Field Optical Microscopy

TL;DR

This paper introduces a deep learning approach to automatically extract optical constants from near-field spectra in scanning near-field optical microscopy, overcoming challenges of nonlinearity and conventional fitting methods.

Contribution

The authors develop a deep learning-based method that improves accuracy, stability, and speed in extracting optical constants compared to traditional fitting techniques.

Findings

Deep learning method outperforms traditional fitting in accuracy.

Enhanced stability against noise in optical parameter extraction.

Faster computational performance in processing near-field spectra.

Abstract

Scanning near-field optical microscopy is one of the most effective techniques for spectroscopy of nanoscale systems. However, inferring optical constants from the measured near-field signal can be challenging because of a complicated and highly nonlinear interaction between the scanned probe and the sample. Conventional fitting methods applied to this problem often suffer from the lack of convergence or require human intervention. Here we develop an alternative approach where the optical parameter extraction is automated by a deep learning network. Compared to its traditional counterparts, our method demonstrates superior accuracy, stability against noise, and computational speed when applied to simulated near-field spectra.