BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:EECS Faculty Candidate Seminar: Guanhong Tao
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20261016T154052Z
UID:tag:localist.com\,2008:EventInstance_45641355802765
DTSTART:20240222T160000Z
DTEND:20240222T170000Z
DESCRIPTION:Towards Secure and Safe AI-enabled Systems Through Optimization
 s\n\nAbstract:\n\nWith the widespread integration of Artificial Intelligen
 ce (AI) in various sectors\, the security and safety of AI-enabled systems
  have not yet been fully ensured. Just like conventional systems having so
 ftware bugs or errors\, applications leveraging AI are not free of bugs. I
 n this talk\, I will present an optimization-based framework for identifyi
 ng and mitigating backdoor vulnerabilities in machine learning models. My 
 talk will cover novel optimization techniques that more efficiently and ef
 fectively detect backdoors in both white-box and black-box settings\, achi
 eving substantial improvement in performance. My work contributed to the P
 urdue team securing the first place in IARPA TrojAI Trojan Detection Compe
 tition (Rounds 1-4). I will share insights on the essence of backdoors and
  their presence in naturally pre-trained models. I will also introduce the
  first hardening framework for mitigating backdoor vulnerabilities. Finall
 y\, I will conclude with an outlook on securing emerging AI techniques\, s
 uch as generative AI (GenAI)\, and the evolving ecosystem enabled by GenAI
 .\n\nBiography:\n\nGuanhong Tao is a PhD candidate at Purdue University\, 
 advised by Prof. Xiangyu Zhang. His research focuses on security and safet
 y of AI-enabled systems. He pinpoints realistic vulnerabilities in real-wo
 rld AI systems and builds practical solutions to mitigating identified vul
 nerabilities and problems using optimizations. He has led the Purdue team 
 to secure the first place in IARPA TrojAI Trojan Detection Competition (Ro
 unds 1-4). He is the recipient of Maurice H. Halstead Memorial Research Aw
 ard in Purdue\, Best Paper Award in ECCV 2022 AROW Workshop\, and CSAW 202
 1 Best Applied Security Paper Award TOP-10 Finalists. His work has been pu
 blished at top-tier security (S&P\, USENIX Security\, CCS\, NDSS)\, machin
 e learning (NeurIPS\, ICML\, ICLR)\, and software engineering (ICSE\, FSE)
  conferences. https://www.cs.purdue.edu/homes/taog/
GEO:35.958178;-83.924516
LOCATION:Min H. Kao Electrical Engineering and Computer Science\, 435
SUMMARY:EECS Faculty Candidate Seminar: Guanhong Tao
URL;VALUE=URI:https://calendar.utk.edu/event/eecs_faculty_candidate_seminar
 _guanhong_tao
CATEGORIES:Lectures & Presentations
END:VEVENT
END:VCALENDAR