BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.earthmonitor.org//8S3XVA
BEGIN:VTIMEZONE
TZID:Europe/Amsterdam
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T020000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-global-workshop-2026-8S3XVA@pretalx.earthmonitor.org
DTSTART;TZID=Europe/Amsterdam:20261009T123000
DTEND;TZID=Europe/Amsterdam:20261009T130000
DESCRIPTION:Satellite Earth Observation (EO) time series are fundamental to
  monitoring our planet's changing environment. However\, inconsistent revi
 sit times and frequent cloud obstruction in optical data (Sentinel-2) ofte
 n force practitioners to rely on lossy data compositing\, which discards c
 ritical phenological information.\nIn this keynote\, we introduce TESSERA 
 (Temporal Embeddings of Surface Spectra for Earth Representation and Analy
 sis)\, a pixel-wise foundation model designed to overcome these challenges
 . TESSERA leverages multi-modal fusion of Sentinel-1 (radar) and Sentinel-
 2 (optical) data\, employing a self-supervised learning framework based on
  Barlow Twins and random temporal sampling. This approach ensures high rob
 ustness to irregular sampling and missing data without requiring expensive
  ground-truth labels.\nA key highlight of TESSERA is its scale and commitm
 ent to Open Science: trained on a global dataset spanning 2017–2025\, th
 e model provides high-dimensional temporal embeddings that capture the "sp
 ectral fingerprint" of the Earth's surface. In alignment with the FAIR pri
 nciples\, we are committed to making TESSERA an open-access resource for t
 he community. We will demonstrate how TESSERA achieves state-of-the-art pe
 rformance in downstream tasks such as crop type mapping and land cover cla
 ssification with minimal labeled data\, paving the way for the next genera
 tion of open-source\, distributed GeoAI monitoring systems.
DTSTAMP:20260624T084615Z
LOCATION:Aula Magna
SUMMARY:TESSERA: A Foundation Model for Label-Efficient and Multi-Modal Ear
 th Observation at Scale - Zhengpeng (Frank) Feng
URL:https://pretalx.earthmonitor.org/global-workshop-2026/talk/8S3XVA/
END:VEVENT
END:VCALENDAR