모델 역전 공격 (Model Inversion Attack)

定义

A privacy attack that reconstructs sensitive training inputs by analyzing a model's outputs.

了解更多

同类别的其他术语