In this work, we will adapt current machine learning (ML) and computer vision (CV) algorithms to develop a suite of scientifically-informed “image search” tools for electron microscopy data. These tools will be used as a means to draw connections between materials images from different sources based on actual quantitative content within the image, and as a mechanism for labelling existing images automatically, using relevant information culled from scientific literature.
|Effective start/end date||12/1/18 → 9/30/20|
- UChicago Argonne, LLC, Argonne National Laboratory (8J-30009-0013A Rev. 0013D)
- Department of Energy (8J-30009-0013A Rev. 0013D)