What is the difference between zero-shot, one-shot, and few-shot learning?

Question

Machine Learning — Hard

What is the difference between zero-shot, one-shot, and few-shot learning?

Accepted Answer

Zero-shot, one-shot, and few-shot learning differ in the number of labeled examples used to adapt to new tasks. Zero-shot requires no examples, one-shot uses one example per class, and few-shot uses a small number of labeled examples per class.