WhiteMail
← 詞彙表

AI 與偵測引擎

적대적 예제 (Adversarial Example)

定義

An input subtly perturbed to be misclassified by a model while appearing normal to humans. A core threat in ML security.

深入了解

相關詞彙