This study explores the application of Convolutional Autoencoders (CAEs) for analyzing and reconstructing Scanning Tunneling Microscopy (STM) images of various crystalline lattice structures. We developed two distinct CAE architectures to process simulated STM images of simple cubic, body-centered cubic (BCC), face-centered cubic (FCC), and hexagonal lattices. Our models were trained on 17×17 pixel patches extracted from 256×256 simulated STM images, incorporating realistic noise characteristics. We evaluated the models' performance using Mean Squared Error (MSE) and Structural Similarity (SSIM) index, and analyzed the learned latent space representations. The results demonstrate the potential of deep learning techniques in STM image analysis, while also highlighting challenges in latent space interpretability and full image reconstruction. This work lays the foundation for future advancements in automated analysis of atomic-scale imaging data, with potential applications in materials science and nanotechnology.
@article{arxiv.2501.13283,
title = {STM Image Analysis using Autoencoders},
author = {Peter Binev and Joshua Moorehead and Ayush Parambath and Luke Parrella and Rori Pumphrey and Miruna Savu},
journal= {arXiv preprint arXiv:2501.13283},
year = {2025}
}