Few-Shot ML GUI Paper Published in Comp Mater Sci

How do we make machine learning more intuitive for autonomous experimentation? Our team’s latest paper in Computational Materials Science describes a unique graphical user interface to classify electron microscope data using few-shot machine learning. Built by Christina Doty and our team of UW-DIRECT capstone students, this work shows the importance of human-computer interaction in AI-driven instrumentation workflows.

From the abstract:

The recent growth in data volumes produced by modern electron microscopes requires rapid, scalable, and flexible approaches to image segmentation and analysis. Few-shot machine learning, which can richly classify images from a handful of user-provided examples, is a promising route to high-throughput analysis. However, current command-line implementations of such approaches can be slow and unintuitive to use, lacking the real-time feedback necessary to perform effective classification. Here we report on the development of a Python-based graphical user interface that enables end users to easily conduct and visualize the output of few-shot learning models. This interface is lightweight and can be hosted locally or on the web, providing the opportunity to reproducibly conduct, share, and crowd-source few-shot analyses.

To view the manuscript, visit: https://doi.org/10.1016/j.commatsci.2021.111121

To download the manuscript directly, click here.