BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//talks.osgeo.org//foss4g-uk-2026//talk//WMRYFL
BEGIN:VTIMEZONE
TZID:GMT
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:GMT
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T020000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:BST
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-foss4g-uk-2026-WMRYFL@talks.osgeo.org
DTSTART;TZID=GMT:20261012T120000
DTEND;TZID=GMT:20261012T123000
DESCRIPTION:Climate-driven disasters like landslides are increasing in freq
 uency\, but the AI models we use to map them are becoming environmentally 
 and financially unsustainable. Modern Earth observation teams face a massi
 ve bottleneck: processing high-resolution topographical data requires imme
 nse\, energy-hungry cloud compute and expensive GPUs.\n\nIn this talk\, I 
 will break down a practical "Green AI" engineering solution to this comput
 e crisis. I will introduce a U-Net++ channel-selection framework\, recentl
 y published in Frontiers in Remote Sensing\, that acts as an intelligent d
 ata filter\, stripping away redundant topographical channels before they e
 ver reach the GPU.\n\nBy optimizing the data pipeline\, this framework suc
 cessfully reduces GPU VRAM requirements by 40% without sacrificing mapping
  accuracy. This approach does more than just lower the hardware barrier fo
 r underfunded disaster response teams\; it actively reduces the carbon foo
 tprint and energy consumption of deploying complex spatial models and also
  makes it faster for a disaster-response scenario.\n\nAttendees will walk 
 away understanding:\nThe environmental and financial toll of unnecessarily
  bloated spatial segmentation models.\nThe architecture behind the VRAM-sa
 ving channel-selection filter.\nHow to apply these techniques to build lea
 ner\, greener geospatial models that can run on accessible\, low-power har
 dware during time-sensitive climate emergencies.
DTSTAMP:20260821T112137Z
LOCATION:Create 1
SUMMARY:Green Spatial AI: Cutting the Compute Cost and time of Climate Disa
 ster Mapping by 40% - Arsalaan Ahmad
URL:https://talks.osgeo.org/foss4g-uk-2026/talk/WMRYFL/
END:VEVENT
END:VCALENDAR