An Acquisition Parameter Study for Machine-Learning-Enabled Electron Backscatter Diffraction

Methods within the domain of artificial intelligence are gaining traction for solving a range materials science objectives, notably use deep neural networks computer vision analysis electron diffraction patterns. An important component deploying these models is an understanding performance as experimental conditions varied. This knowledge can inspire confidence in classifications over operating and identify where degraded. Elucidating relative impact each parameter will suggest most parameters to vary during collection future training data. Knowing which data efforts prioritize concern given time required collect or simulate vast libraries patterns wide variety without considering varying any parameters. In this work, five parameters, frame averaging, detector tilt, sample-to-detector distance, accelerating voltage, pattern resolution, essential individually varied backscatter explore effect on produced by network trained from captured using fixed set The model shown be resilient nearly all individual changes examined here.