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:20260730T152638Z
LOCATION:
DTSTART;TZID=America/Los_Angeles:20260728T125200
DTEND;TZID=America/Los_Angeles:20260728T125300
UID:dac_DAC 2026_sess315_RESEARCH1960@linklings.com
SUMMARY:HyperLiDAR: Adaptive Post-Deployment LiDAR Segmentation via Hyperd
 imensional Computing
DESCRIPTION:Ivannia Gomez Moreno, Yi Yao, and Ye Tian (University of Calif
 ornia, San Diego); Xiaofan Yu (University of California Merced); Flavio Po
 nzina (San Diego State University); Michael Sullivan and Jingyi Zhang (Uni
 versity of California, San Diego); Mingyu Yang and Hun Seok Kim (Universit
 y of Michigan); and Tajana Rosing (University of California, San Diego)\n\
 nLiDAR semantic segmentation plays a pivotal role in scene understanding f
 or edge applications such as autonomous driving. Once deployed in the real
  world, the model must adapt to its surrounding environment through rapid 
 adaptive training and updates, even with limited compute and energy resour
 ces\non an edge device. Existing segmentation models rely on large neural 
 networks, which need significant memory and computing resources for post-d
 eployment adaptation. To address the above challenges, we introduce HyperL
 iDAR, the first lightweight LiDAR segmentation framework based on Hyperdim
 ensional\nComputing (HDC) for adapting to streaming point cloud scans afte
 r deployment. HDC is a brain-inspired approach well-suited for efficient o
 n-device learning. HyperLiDAR combines a pretrained feature extractor with
  HDC training to support lightweight adaptation on edge devices. We furthe
 r design a buffer selection\nstrategy to mitigate the high data volume in 
 each scan. Extensive evaluations on two state-of-the-art LiDAR segmentatio
 n benchmarks and three representative devices demonstrate that HyperLiDAR 
 outperforms state-of-the-art segmentation methods and accelerates retraini
 ng speed by up to 13.8×.\n\nTopics: Quantum\n\nTrack: DES6. Quantum Comput
 ing, Systems\n\nSession Chairs: Hokeun Kim (Arizona State University); Bae
 kGyu Kim (Daegu Gyeongbuk Institute of Science and Technology); Xiang Chen
  (Peking University); Sangyeob Kim (Yonsei University); Robert M. Radway (
 University of Pennsylvania); Prashanth Mohan (University of Pittsburgh); X
 iaoxuan Yang (University of Virginia, Stanford University); Ramtin Zand (U
 niversity of South Carolina); Hadi Kamali (University of Central Florida);
  Jiafeng Xie (Villanova University); Prabuddha Chakraborty (University of 
 Maine); Junichiro Kadomoto (University of Tokyo); and Yosuke Ueno (RIKEN)\
 n\n
END:VEVENT
END:VCALENDAR