Attribute-Based Learning

Zero-Shot Learning (ZSL)

Machine learning paradigm where the model recognizes classes never seen during training by using auxiliary semantic information such as textual descriptions or attributes.

← Zurück