Explore a 19-minute IEEE conference talk that delves into the concept of provable robustness against backdoor attacks in machine learning systems. Learn about the innovative RAB (Robust Against Backdoor) approach presented by researchers from ETH Zurich and the University of Illinois at Urbana-Champaign. Gain insights into the challenges posed by backdoor attacks and discover how the proposed method aims to provide provable guarantees for model robustness. Understand the implications of this research for enhancing the security and reliability of machine learning models in various applications.
Overview
Syllabus
RAB: Provable Robustness Against Backdoor Attacks
Taught by
IEEE Symposium on Security and Privacy
Tags
Related Courses
-
BadEncoder - Backdoor Attacks to Pre-trained Encoders in Self-Supervised Learning
-
SoK - Certified Robustness for Deep Neural Networks
-
Jigsaw Puzzle - Selective Backdoor Attack to Subvert Malware Classifiers
-
RoFL - Robustness of Secure Federated Learning
-
Selective Amnesia - On Efficient, High-Fidelity and Blind Suppression of Backdoor Effects in Trojaned Machine Learning Models
-
Instruction Backdoor Attacks Against Customized Large Language Models
