Neural RHEED alignment with limited training data during CdTe MBE growth

We introduce a data-efficient neural-vision assisted method to automate crystallographic alignment during molecular beam epitaxy (MBE) growth.

We introduce a data-efficient neural-vision assisted method to automate crystallographic alignment during molecular beam epitaxy (MBE) growth. Trained on reflection high-energy electron diffraction (RHEED) patterns from only 15 CdTe structures, our model - enabled by…