Microstructural Hierarchy Descriptor Enabling Interpretative AI for Microelectronic Failure Analysis

Abstract This article proposes the MicroStructural Hierarchy Descriptor (µSHD) as a systematic and quantitative approach to spectra and image data in microelectronics failure analysis. It discusses concrete routes for employing µSHD directly as the quantitative descriptor for supervised and unsupervised machine learning. The authors propose that µSHD tools can be used to automate and improve characterization techniques and image processing and analysis protocols.