-
A Survey on Evaluation Metrics for Synthetic Material Micro-Structure Images from Generative Models
Authors:
Devesh Shah,
Anirudh Suresh,
Alemayehu Admasu,
Devesh Upadhyay,
Kalyanmoy Deb
Abstract:
The evaluation of synthetic micro-structure images is an emerging problem as machine learning and materials science research have evolved together. Typical state of the art methods in evaluating synthetic images from generative models have relied on the Fréchet Inception Distance. However, this and other similar methods, are limited in the materials domain due to both the unique features that char…
▽ More
The evaluation of synthetic micro-structure images is an emerging problem as machine learning and materials science research have evolved together. Typical state of the art methods in evaluating synthetic images from generative models have relied on the Fréchet Inception Distance. However, this and other similar methods, are limited in the materials domain due to both the unique features that characterize physically accurate micro-structures and limited dataset sizes. In this study we evaluate a variety of methods on scanning electron microscope (SEM) images of graphene-reinforced polyurethane foams. The primary objective of this paper is to report our findings with regards to the shortcomings of existing methods so as to encourage the machine learning community to consider enhancements in metrics for assessing quality of synthetic images in the material science domain.
△ Less
Submitted 3 November, 2022;
originally announced November 2022.
-
The Laser-Diode Heated Floating Zone Method for Automated Optimal Synthesis of Refractory Oxides and Alloys
Authors:
Alemayehu S. Admasu,
Durga Sankar Vavilapalli
Abstract:
We perform an experimental investigation of the synthesis of complex materials using the Laser-diode heated floating zone method (L-FZ). We will briefly introduce the Infrared-heated floating zone method of bulk crystal growth and then delve into recent advances in using a Laser-diode heated floating zone method. We demonstrate L-FZ for the growth of large high-quality single crystalline samples o…
▽ More
We perform an experimental investigation of the synthesis of complex materials using the Laser-diode heated floating zone method (L-FZ). We will briefly introduce the Infrared-heated floating zone method of bulk crystal growth and then delve into recent advances in using a Laser-diode heated floating zone method. We demonstrate L-FZ for the growth of large high-quality single crystalline samples of of the n=2 Ruddlesden-Popper homologous series $RE_2SrAl_2O_7$, RE = Nd, Sm, Eu, Gd, Tb, Dy with physically interesting properties (e.g spin-ice) as well as the incongruently melting stuffed-tridymite type oxide $BaCoSiO_4$. We conclude with a summary of the results and future research directions in automating crystal growth, which will open access into the synthesis and characterization of previously unstudied class of materials such as refractory complex oxides and alloys.
△ Less
Submitted 9 March, 2021;
originally announced March 2021.
-
Image registration of low signal-to-noise cryo-STEM data
Authors:
Benjamin H. Savitzky,
Ismail El Baggari,
Colin Clement,
Emily Waite,
John P. Sheckelton,
Christopher Pasco,
Alemayehu S. Admasu,
Jaewook Kim,
Sang-Wook Cheong,
Tyrel M. McQueen,
Robert Hovden,
Lena F. Kourkoutis
Abstract:
Combining multiple fast image acquisitions to mitigate scan noise and drift artifacts has proven essential for picometer precision, quantitative analysis of atomic resolution scanning transmission electron microscopy (STEM) data. For very low signal-to-noise ratio (SNR) image stacks - frequently required for undistorted imaging at liquid nitrogen temperatures - image registration is particularly d…
▽ More
Combining multiple fast image acquisitions to mitigate scan noise and drift artifacts has proven essential for picometer precision, quantitative analysis of atomic resolution scanning transmission electron microscopy (STEM) data. For very low signal-to-noise ratio (SNR) image stacks - frequently required for undistorted imaging at liquid nitrogen temperatures - image registration is particularly delicate, and standard approaches may either fail, or produce subtly specious reconstructed lattice images. We present an approach which effectively registers and averages image stacks which are challenging due to their low-SNR and propensity for unit cell misalignments. Registering all possible image pairs in a multi-image stack leads to significant information surplus. In combination with a simple physical picture of stage drift, this enables identification of incorrect image registrations, and determination of the optimal image shifts from the complete set of relative shifts. We demonstrate the effectiveness of our approach on experimental, cryogenic STEM datasets, highlighting subtle artifacts endemic to low-SNR lattice images and how they can be avoided. High-SNR average images with information transfer out to 0.72 A are achieved at 300 kV and with the sample cooled to near liquid nitrogen temperature.
△ Less
Submitted 23 October, 2017;
originally announced October 2017.
