[Colloquium] Reminder: Li/MS Presentation/Nov 22, 2019

[Margaret Jaffey]

This is a reminder about Huiying Li's MS Presentation tomorrow.

------------------------------------------------------------------------------
Date:  Friday, November 22, 2019

Time:  12:30 PM

Place:  John Crerar Library 354

M.S. Candidate:  Huiying Li

M.S. Paper Title: LATENT BACKDOOR ATTACKS ON DEEP NEURAL NETWORKS

Abstract:
Backdoor attacks on deep neural networks (DNNs) are hidden malicious
behaviors embedded into DNN models, where misclassification rules are
hidden inside normal models, only to be triggered by very specific
inputs. When models are compromised, extremely dangerous consequences
might happen since DNNs are widely deployed in safety and
security-critical areas like self-driving cars. However, these
traditional backdoors assume a context where users train their own
models from scratch, which rarely occurs in practice. Instead, users
typically customize Teacher models which are pretrained by model
providers like Google, through a process called transfer learning.
This customization process introduces significant changes to models
and disrupts hidden backdoors, greatly reducing the actual impact of
backdoors in practice. In this study, we describe latent backdoors, a
more powerful and stealthy variant of backdoor attacks that functions
under transfer learning. Latent backdoors are incomplete backdoors
embedded into a Teacher model, and automatically inherited by multiple
Student models through transfer learning. If any Student models
include the label targeted by the backdoor, then its customization
process completes the backdoor and makes it active. We show that
latent backdoors can be quite effective in a variety of application
contexts, and validate its practicality through real-world attacks
against traffic sign recognition, iris identification of volunteers,
and facial recognition of public figures (politicians). Finally, we
evaluate 4 potential defenses, and find that only one is effective in
disrupting latent backdoors, but might incur a cost in classification
accuracy as tradeoff.

Huiying's advisors are Prof. Ben Zhao and Prof. Heather Zheng

Login to the Computer Science Department website for details:
 https://newtraell.cs.uchicago.edu/phd/ms_announcements#huiyingli

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Margaret P. Jaffey            margaret at cs.uchicago.edu
Department of Computer Science
Student Support Rep (Ry 156)               (773) 702-6011
The University of Chicago      http://www.cs.uchicago.edu
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=