BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20241120T082408Z
LOCATION:HG F 1
DTSTART;TZID=Europe/Stockholm:20240603T120000
DTEND;TZID=Europe/Stockholm:20240603T123000
UID:submissions.pasc-conference.org_PASC24_sess108_msa174@linklings.com
SUMMARY:Foundation Models in Earth Science and Remote Sensing
DESCRIPTION:Minisymposium\n\nJohannes Jakubik (IBM Research)\n\nSignifican
 t progress in the development of highly adaptable and reusable Artificial 
 Intelligence (AI) models is expected to have a significant impact on Earth
  science and remote sensing. Foundation Models (FMs), AI models designed t
 o replace task-specific models, are increasingly being recognized for thei
 r versatility across numerous downstream applications. These models, train
 ed using self-supervised techniques on any type of sequence data, circumve
 nt the need for large annotated datasets, a major bottleneck in traditiona
 l AI model development. FMs can be applied to downstream tasks using few-s
 hot learning and fine-tuning, significantly reducing the need for large la
 beled training datasets and computational resources. In contrast to task-s
 pecific models, large-scale FMs facilitate the processing of multi-modal d
 ata from different satellites and additional data modalities in order to o
 btain improved model skills in the Earth observation domain.\n\nDomain: Cl
 imate, Weather, and Earth Sciences, Computational Methods and Applied Math
 ematics\n\nSession Chair: Christian Lessig (ECMWF, Otto-von-Guericke-Unive
 rsitat Magdeburg)
END:VEVENT
END:VCALENDAR