Adversarial Attacks Explained
Katharine discusses the intricacies of adversarial attacks in machine learning, particularly focusing on targeted and untargeted approaches. She explains how attackers can manipulate models without needing access to the original, utilizing black box methods to achieve their goals. The conversation highlights the challenges in creating effective adversarial examples, especially when distinguishing between complex objects.In this clip
From this podcast
Software Engineering Radio - the podcast for professional software developers
Episode 395: Katharine Jarmul on Security and Privacy in Machine Learning
Related Questions