Noun
few-shot learning (uncountable)
(machine learning) An object categorization problem, mainly in computer vision, involving the classification of objects based on only a few examples.
Coordinate terms: N-shot learning, one-shot learning, zero-shot learning