Defending Against Adversarial Model Attacks Using Kubeflow
Offered By: Linux Foundation via YouTube
Course Description
Overview
Explore defensive strategies against adversarial model attacks in this 43-minute conference talk by IBM experts Animesh Singh and Andrew Butler. Delve into the vision for Trusted AI, understand various adversarial threats, and learn how to leverage Kubeflow and Kubeflow pipelines for robust model protection. Gain insights into the Square Attack technique, examine key parameters, and witness practical demonstrations of explainer flows. Discover the importance of building adversarial teams and acquire valuable contact information for further inquiries on enhancing AI model security.
Syllabus
Introduction
Presentation
Vision for Trusted AI
Adversarial Threats
Art and Practice
Kubeflow
Kubeflow pipelines
Square Attack
Parameters
Explainer Spec
Explainer Flow
Explainer Demo
Adversarial Team
Contact Information
Conclusion
Taught by
Linux Foundation
Tags
Related Courses
TinyML Talks - Software-Hardware Co-design for Tiny AI SystemstinyML via YouTube Cross-Domain Transferability of Adversarial Perturbations - CAP6412 Spring 2021
University of Central Florida via YouTube InfoSec Deep Learning in Action
nullcon via YouTube Zen and the Art of Adversarial Machine Learning
Black Hat via YouTube Practical Defenses Against Adversarial Machine Learning
Black Hat via YouTube