摘要：
Basic iterative method(BIM):论文地址 笔记地址 Projected gradient descent(PGD):论文地址 笔记地址 区别1 来自于：https://www.sciencedirect.com/science/article/pii/S20958099193 阅读全文
【论文笔记】（2015，防御蒸馏）Distillation as a Defense to Adversarial Perturbations against Deep Neural Networks
摘要：
有关蒸馏 (Distillation)的论文： (2006)Model Compression (2014)Do Deep Nets Really Need to be Deep? 论文笔记 (2015)Distilling the Knowledge in a Neural Network 论文笔 阅读全文