【Google】深度学习对抗鲁棒性，43页ppt

最新深度学习对抗鲁棒性教程

• 深度学习基础 Deep learning essentials
• 对抗扰动 Introduction to adversarial perturbations
  • Simple Projected Gradient Descent-based attacks
  • Targeted Projected Gradient Descent-based attacks
  • Fast Gradient Sign Method (FGSM) attacks
  • Natural [8]
  • Synthetic [1, 2]
• Optimizer susceptibility w.r.t to different attacks 优化器对不同攻击的敏感性w.r.
• 对抗学习 Adversarial learning
  • Training on a dataset perturbed with FGSM
  • Training with Neural Structured Learning [3]
• Improving adversarial performance with EfficientNet [4] and its variants like Noisy Student Training [5] and AdvProp [6]

https://github.com/dipanjanS/adversarial-learning-robustness