Zero-Shot Learning (ZSL)

What is Zero-Shot Learning (ZSL)?

The ability of a model to recognize classes it has never seen during training, often by using auxiliary information or semantic relationships.

Where did the term "Zero-Shot Learning (ZSL)" come from?

Solving the long-tail problem.

How is "Zero-Shot Learning (ZSL)" used today?

Critical for generalized AI systems like CLIP.

Related Terms

• Machine Learning
• Transfer Learning