[Colloquium] Today: Hadleigh Schwartz MS Presentation/May 23, 2022

[Jessica Garza]

This is a reminder about Hadleigh Schwartz’s MS presentation.



Hadleigh Schwartz is a student in the Bx/MS program.

—————————————————————————————————————————————

Date: Monday, May 23, 2022

Time: 2 PM, CST

Location: JCL 298

M.S. Candidate: Hadleigh Schwartz

M.S. Paper Title: Towards Systematic Evaluation of Privacy Risks in IoT Video Analytics

Advisor: Prof. Heather Zheng

Committee Members: Prof. Heather Zheng , Prof. Ben Zhao, and Prof. Junchen Jiang

—————————————————————————————————————————————

Abstract:

Today’s video analytics systems can parse scenes and human activity from videos collected by IoT cameras and are increasingly being deployed for critical tasks such as security and traffic management. Yet their ability to continuously monitor and identify individuals also raises significant privacy concerns.  In this work, we propose a method for systematically evaluating the privacy risks of video analytics using real world IoT surveillance video. Our method takes into account policy, edge computation, and camera placement constraints to explore possible configurations and risks.  Our evaluation shows that gait recognition is a significant privacy threat that should be considered alongside facial recognition and accounted for in modern privacy policies. Furthermore, we find edge-based verification methods still perform significantly worse than those methods requiring cloud computing. Lastly, we find evidence that color recognition, which we use to identify individuals' shirt colors, can be a helpful modality for increasing identification confidence and efficiency. We believe that the approach we present provides a starting point for reasoning about privacy risks of IoT video analytics under various policy and computation constraints, in both the present and future.

—————————————————————————————————————————————





Jessica Garza
Assistant Director of Undergraduate Studies
Department of Computer Science
The University of Chicago
John Crerar Library 374
Office: (773) 702-2336



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20220523/9815bacd/attachment-0002.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: MS Thesis - Hadleigh Schwartz.pdf
Type: application/pdf
Size: 4795974 bytes
Desc: not available
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20220523/9815bacd/attachment-0001.pdf>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20220523/9815bacd/attachment-0003.html>