TerraBytes – Towards global datasets and models for Earth Observation

19 July, 2025
Vancouver, Canada
ICML 2025 Workshop

 

TerraBytes – Towards global datasets and models for Earth Observation is an official Workshop of the ICML 2025 Forty-Second International Conference on Machine Learning.

Event website

Earth Observation (EO) presents unique challenges and opportunities that set it apart from other fields of machine learning (ML) and computer vision (CV) (Rolf et al., 2024). EO data is abundant, repeatedly covering a large but limited environment: our planet. The colocation and evolution of these observations is a rich, emergent source of information, of multimodal and multitemporal nature. However, due to both local and global changes, especially climate change, the statistical distribution of EO data is inherently non-stationary (Tarasiou et al., 2023). These properties break some of the usual assumptions of ML, and EO data require special care for data handling and modeling (Mai et al., 2022). In addition, current EO datasets are sampled with spatio-temporal biases.

Some areas, e.g., the global South, are strongly under-represented within EO datasets (Cornebise et al., 2022). In optical imagery, cloud cover is undesirable, thus leading to datasets that remove cloudy images at the risk of biasing their geographical coverage (Tiede et al., 2021). Addressing these distributional biases is of primary importance, as they have an impact on the performance and reliability of models for downstream applications in ecology, geosciences, agriculture, urban planning, etc. (Kattenborn et al., 2022).

TerraBytes is an initiative to address these challenges. At the intersection of data curation, data archiving, and representation learning, this workshop will foster a holistic discussion covering major steps in the EO from downlinked satellite data, training paradigms to downstream applications.

---

Program

Morning session

• 9:00-9:10: Opening remarks
• 9:10-9:50: Keynote: Sujit Roy (NASA IMPACT)
• 9:50-10:10: Paper spotlights: Benchmarks and Deployments
• Coffe break
• 10:40-11:20: Keynote: Razvan Cosac (ESA Ground Segment) “From TerraBytes to PetaBytes: ESA Earth Observation Framework and Copernicus Data Ecosystem”
• 11:20-12:00: Paper spotlights: Multimodality

Afternoon session

• 13:15-14:00: Paper spotlights: New Methods
• 14:00-15:00: Poster session
• Coffee break
• 15:30-16:10: Keynote: Erin Trochim (University of Alaska Fairbanks) “Building Trust in GeoAI and Earth Observations: Lessons from Complex Systems”
• 16:10-17:10: Panel session: “Towards Global Geospatial Models”
• 17:10-17:20: European Space Agency Copernicus AI Report: Stakeholder Feedback Session
• 17:20-17:30: Closing remarks

---

Speakers

• Sujit Roy (NASA IMPACT)
• Razvan Cosac (ESA Ground Segment)
• Erin Trochim (University of Alaska Fairbanks)

---

Organizers

• Nicolas Audebert, LASTIG, IGN
• Hossein Azizpour, KTH Royal Institute of Technology
• Valentin Barriere, University of Chile
• Javiera Castillo Navarro, CNAM
• Mikolaj Czerkawski, Asterisk Labs
• Heng Fang, KTH Royal Institute of Technology
• Alistair Francis, Asterisk Labs
• Valerio Marsocci, ESA Phi Lab
• Andrea Nascetti, KTH Royal Institute of Technology
• Ritu Yadav, KTH Royal Institute of Technology

---

Sponsors

• CopernicusLAC Chile
• Image Analysis and Data Fusion Technical Committee (IADF TC) of the IEEE Geoscience and Remote Sensing Society

This event received funding from Proyecto Basal ANID FB210005.

Posted on Jul 4, 2025 in Workshops & Congresses