A few-shot learning workflow segments and classifies STM defects on Si, Ge, and TiO2, reaching 93% 1-shot accuracy on silicon but only 61-70% on the other surfaces.
Choudhary, Camp C
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cond-mat.mtrl-sci 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Overcoming Labelled Data Scarcity for Defect Classification in Scanning Tunneling Microscopy
A few-shot learning workflow segments and classifies STM defects on Si, Ge, and TiO2, reaching 93% 1-shot accuracy on silicon but only 61-70% on the other surfaces.