BriefGPT.xyz
Jun, 2023
克服基于敌对攻击的人在环应用
Overcoming Adversarial Attacks for Human-in-the-Loop Applications
HTML
PDF
Ryan McCoppin, Marla Kennedy, Platon Lukyanenko, Sean Kennedy
TL;DR
人类分析对深度神经网络的鲁棒性产生积极影响，尚未在对抗机器学习文献中得到很好的探究。人类视觉注意力模型可能会提高人-机图像分析系统的解释性和鲁棒性。虽然存在挑战，但需要进一步研究，以便选择适宜的可视化解释，以便图像分析员评估所给数据模型。
Abstract
Including human analysis has the potential to positively affect the robustness of
deep neural networks
and is relatively unexplored in the
adversarial machine learning
literature.
→