few-shot learning

Noun

 * 1)  An object categorization problem, mainly in computer vision, involving the classification of objects based on only a few examples.