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:20240626T180034Z
LOCATION:3008\, 3rd Floor
DTSTART;TZID=America/Los_Angeles:20240626T113000
DTEND;TZID=America/Los_Angeles:20240626T114500
UID:dac_DAC 2024_sess126_RESEARCH1535@linklings.com
SUMMARY:HiRISE: High-Resolution Image Scaling for Edge ML via In-Sensor Co
 mpression and Selective ROI
DESCRIPTION:Research Manuscript\n\nBrendan Reidy (University of South Caro
 lina); Sepehr Tabrizchi (University of Nebraska, Lincoln); MohammadReza Mo
 hammadi (University of South Carolina); Shaahin Angizi (New Jersey Institu
 te of Technology); Arman Roohi (University of Nebraska, Lincoln); and Ramt
 in Zand (University of South Carolina)\n\nWith the rise of tiny IoT device
 s powered by machine learning (ML), many researchers have directed their f
 ocus toward compressing models to fit on tiny edge devices. Recent works h
 ave achieved remarkable success in compressing ML models for object detect
 ion and image classification on microcontrollers with small memory, e.g., 
 512kB SRAM. However, there remain many challenges prohibiting the deployme
 nt of ML systems that require high-resolution images. Due to fundamental l
 imits in memory capacity for tiny IoT devices, it may be physically imposs
 ible to store large images without external hardware. To this end, we prop
 ose a high-resolution image scaling system for edge ML, called HiRISE, whi
 ch is equipped with selective region-of-interest (ROI) capability leveragi
 ng analog in-sensor image scaling. Our methodology not only significantly 
 reduces the peak memory requirements, but also achieves up to 17.7x reduct
 ion in data transfer and energy consumption.\n\nTopic: AI, Design\n\nKeywo
 rd: AI/ML System and Platform Design\n\nSession Chair: Hsien-Hsin Sean Lee
  (Intel Corporation)
END:VEVENT
END:VCALENDAR