BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250625T183020Z
LOCATION:Level 2 Lobby
DTSTART;TZID=America/Los_Angeles:20250623T180000
DTEND;TZID=America/Los_Angeles:20250623T190000
UID:dac_DAC 2025_sess262_RESEARCH1081@linklings.com
SUMMARY:Non-Negative AdderNet (NNAN): Can We Make DNNs More Secure and Eff
 icient Without Multiplication?
DESCRIPTION:Yunxiang Zhang, Sabbir Ahmed, Abeer Almalky, Adnan Siraj Rakin
 , and Wenfeng Zhao (Binghamton University)\n\nIn the vision domain, AdderN
 et emerges as a hardware-efficient alternative to traditional Convolutiona
 l Neural Networks (CNNs) by replacing multiplications with additions. Howe
 ver, the security aspect of AdderNet remains under-explored. To this end, 
 we consider a pivotal question: Does AdderNet compromise model security to
  hardware efficiency. In this paper, we extend the investigation on the Ad
 derNet vulnerabilities to adversarial weight perturbation attack, e.g., Bi
 t-Flip Attack (BFA) for the first time, and empirically demonstrate that A
 dderNet is indeed more susceptible against BFA. To preserve both the hardw
 are efficiency of AdderNet and defend against BFA, we propose a novel Secu
 re Non-Negative AdderNet (NNAN) model incorporating a lightweight non-posi
 tive weight encoding technique. NNAN enables lightweight, real-time detect
 ion and correction by securing (a) the Most Significant Bits (MSBs) of all
  weights and (b) second MSBs in the non-positive encoded weights. To test 
 the resilience of our proposed defense further, we perform an advanced BFA
  (BFA+) where the attacker has sufficient knowledge about the proposed enc
 oding scheme. Our results indicate that NNAN defends BFA successfully and 
 requires a higher number of attack cycles to be compromised against BFA+, 
 which concludes robustness can be achieved through additive architectures 
 w/o multiplication.\n\n
END:VEVENT
END:VCALENDAR
