AtomVision: A Machine Vision Library for Atomistic Images

Abstract

Computer vision techniques have immense potential for materials design applications. In this work, we introduce an integrated and general-purpose AtomVision library that can be used to generate and curate microscopy image (such as scanning tunneling microscopy and scanning transmission electron microscopy) data sets and apply a variety of machine learning techniques. To demonstrate the applicability of this library, we (1) establish an atomistic image data set of about 10 000 materials with large structural and chemical diversity, (2) develop and compare convolutional and atomistic line graph neural network models to classify the Bravais lattices, (3) demonstrate the application of fully convolutional neural networks using U-Net architecture to pixelwise classify atom versus background, (4) use a generative adversarial network for super resolution, (5) curate an image data set on the basis of natural language processing using an open-access arXiv data set, and (6) integrate the computational framework with experimental microscopy images for Rh, Fe 3 O 4 , and SnS systems. The AtomVision library is available at https://github.com/ usnistgov/atomvision.

ChemNLP: A Natural Language-Processing-Based Library for Materials Chemistry Text Data

“…The models and data will also be integrated in a larger benchmarking project called JARVIS-Leaderboard () for enhancing reproducibility and transparency. Moreover, we plan to extend ChemNLP to generate multimodality projects by integrating with other projects in JARVIS such as atomistic vision (AtomVision) and atomistic line graph neural network (ALIGNN) libraries in the future.…”

InterMat: accelerating band offset prediction in semiconductor interfaces with DFT and deep learning

“…50,51 Such models have oen been applied for bulk property predictions and their applicability for defects and interfaces remains an open question. Several machine learning tools available in JARVIS such as classical force-eld inspired descriptors (CFID), 43 atomistic line graph neural network (ALIGNN), 52,53 computer vision for atomistic images (Atom-Vision) 54 and natural language processing for chemistry (ChemNLP) 55,56 can be used in this regards to accelerate the interface design tasks. In particular, ALIGNN has been used to develop several fast surrogate models for property predictions as well as a unied force-eld for fast structure optimizations.…”

Automated Generation and Analysis of Molecular Images Using Generative Artificial Intelligence Models

“…More specifically, a large database of molecular/atomic scanning tunneling microscopy (STM) images is absent. Prohibited by the sophisticated experimentations and the intensive labor for acquiring high-resolution molecular images, most data-driven research involving high-resolution SPM images builds on specialized ML models or rely on simulated images. − In this work, we have developed an ML model based on an experimental data set specifically tailored for high-fidelity simulation of molecular STM imaging. Our model effectively captures the subtle variations in STM images resulting from diverse molecular conformations and orientations.…”