科研篇四：對抗樣本20篇-ICML2019

文章目錄 一、對抗攻擊（Adversarial Attack） 1.1.Adversarial Attacks on Node Embeddings via Graph Poisoning 1.2.Adversarial camera stickers: A physical camera-based attack on deep learning systems 二、對抗防禦 2.1.Using

>>阅读原文<<