We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Computational Physics

Title: Deep Bayesian Local Crystallography

Abstract: The advent of high-resolution electron and scanning probe microscopy imaging has opened the floodgates for acquiring atomically resolved images of bulk materials, 2D materials, and surfaces. This plethora of data contains an immense volume of information on materials structures, structural distortions, and physical functionalities. Harnessing this knowledge regarding local physical phenomena necessitates the development of the mathematical frameworks for extraction of relevant information. However, the analysis of atomically resolved images is often based on the adaptation of concepts from macroscopic physics, notably translational and point group symmetries and symmetry lowering phenomena. Here, we explore the bottom-up definition of structural units and symmetry in atomically resolved data using a Bayesian framework. We demonstrate the need for a Bayesian definition of symmetry using a simple toy model and demonstrate how this definition can be extended to the experimental data using deep learning networks in a Bayesian setting, namely rotationally invariant variational autoencoders.
Comments: Combined Paper and Supplementary Information. 40 pages. 8 Figures plus 12 Supplementary figures
Subjects: Computational Physics (physics.comp-ph)
Cite as: arXiv:2012.07134 [physics.comp-ph]
  (or arXiv:2012.07134v1 [physics.comp-ph] for this version)

Submission history

From: Mark Oxley [view email]
[v1] Sun, 13 Dec 2020 20:01:31 GMT (9697kb)

Link back to: arXiv, form interface, contact.