Introduction to Fundamentals, Applications and Libraries of Adversarial Attacks

AAISS

Syllabus

• Machine Learning and its significance in today’s life and therefore its vulnerability in sensitive tasks
• The potential of adversarial attacks and some examples
• Basics of Adversarial attacks
• Well-known adversarial attacks
  - Libraries and some sample codes for the implementation of adversarial attacks
• Defending Machine Learning algorithms against adversarial attacks
• Adversarial attacks in voice recognition, image processing and cybersecurity

Papers

FGSM: https://arxiv.org/abs/1412.6572
DeepFool: https://arxiv.org/abs/1511.04599
Boundary Attack: https://arxiv.org/abs/1712.04248
Audio Adversarial Examples: https://arxiv.org/abs/1801.01944
Ensemble Adversarial Training: https://arxiv.org/abs/1705.07204
Defensive Distillation: https://arxiv.org/abs/1511.04508
Stateful Defense: https://arxiv.org/abs/1907.05587

Github

Audio Adversarial Examples: https://github.com/carlini/audio_adversarial_examples
Defensive Distillation: https://github.com/carlini/nn_robust_attacks
Stateful Detection: https://github.com/schoyc/blackbox-detection

Thanks to Bardia Esmaeili.