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:20260417T190156Z
LOCATION:West Building\, Rooms 301-305
DTSTART;TZID=America/Los_Angeles:20250814T080000
DTEND;TZID=America/Los_Angeles:20250814T084500
UID:siggraph_SIGGRAPH 2025_sess594_ftalk_105@linklings.com
SUMMARY:Talk: Earth Embeddings: Harnessing the Information in Earth Observ
 ation Data with Machine Learning
DESCRIPTION:Esther Rolf (University of Colorado Boulder)\n\nMachine learni
 ng (ML) for Earth Observation (EO) data is revolutionizing the speed and s
 cope at which science and policy can operate — filling critical data gaps 
 across fields such as ecology and development economics. In this talk, I w
 ill outline a class of ML for EO models that distill global satellite data
  into compact, multi-purpose representations of the Earth. I’ll trace the 
 recent evolution of these “Earth embedding” models, from early image embed
 dings designed to capture the unique characteristics of satellite imagery,
  to an emerging class of location encoders that serve as implicit neural r
 epresentations of EO data. After discussing the impact-driven goals and me
 thodological details of these models, I'll conclude by discussing my longe
 r term vision of building Earth embedding models to unlock new scales of s
 cience.\n\nInterest Area: Research & Education\n\nRecording: Livestreamed,
  Recorded\n\nKeyword: Artificial Intelligence/Machine Learning, Computer V
 ision, Digital Twins\n\nRegistration Category: Full Conference, Virtual Ac
 cess, Experience\n\n
END:VEVENT
END:VCALENDAR