A Deep Dive into FSHA Privacy Reconstruction Attacks

date:

Tuesday, Oct 4, 2022

Time:

7:43 am

Summary:

How can we enable privacy preserving deep learning on private data silos? This talk explores the vulnerabilities of a promising solution: split learning enhanced with differential privacy. We focus on FSHA (Feature Space Hijacking Attacks) reconstruction attacks using GAN. We describe the attacker’s neural network architecture. We follow with constructing and running an attack on distributed data silos. Finally, we analyse the successfully reconstructed private inputs and conclude with possible attack risk mitigations.

Ready to attend?

Register now! Join your peers.

Register now View agenda
Newsletter

Knowledge is everything!
Sign up for our newsletter to receive:

  • 10% off your ticket!
  • insights, interviews, tips, news, and much more about Deep Learning World
  • price break reminders