Workshop Program
1 October 2026 | Room B
| Time | Program Details | Keynote Information | Paper Information |
|---|---|---|---|
| 9:00 - 10:31 | • Keynote 1 (45 min + 10 min Q&A) • Paper presentations (3 × 10 min + 2 min Q&A) | Damian Borth University of St.Gallen GeoSANE: Learning Geospatial Representations from Models, Not Data | Papers: • Optimizing Onboard Earth Observation AI for Robustness against Sensor Degradations (Ivar van der Spoel, Lorenzo Papa, Roberto Del Prete, Andre Deutz, Gabriele Meoni, Jan van Rijn) • Credibility-Aware AutoML for Small Earth-Observation Constellations: The Two-to-Three-Satellite Handoff as a Search Objective (Agata Wijata, Przemysław Biecek) • Design Choice Sensitivity in Remote Sensing Multi-Label Classification (Maryam Gholami Shiri, Eva Tuba, Sašo Džeroski, Tome Eftimov, Ana Nikolikj) |
| 10:31 - 11:00 | Coffee Break | ||
| 11:00 - 12:40 | • Keynote 2 (30 min + 10 min Q&A) • Paper presentations (5 × 10 min + 2 min Q&A) | Mitra Baratchi Leiden University AutoML to Advance Earth Observation Research | Papers: • Automated Machine Learning for Sea Ice Concentration Charting (Sven van Collenburg, Andreas Stokholm, Holger Hoos, Jan van Rijn) • Insights from Autoresearch for Solar Panel Segmentation (Justinas Lekavičius, Kürşat Kömürcü, Valentas Gružauskas, Linas Petkevičius) • Spatial Cross-Validation for Machine Learning Systems (Abygaïl Stegenga, Julia Wąsala, Jan Van Rijn, Mitra Baratchi) • AI for Materials Science: Tuning Laser-Induced Graphene Production (Lars Kotthoff) • Evolutionary Architecture Search for Chlorophyll-𝑎 Prediction in Lakes using Sentinel-2 (Kürşat Kömürcü, Linas Petkevicius) |
| 12:40 - 13:30 | Lunch | ||
| 13:30 - 15:08 | • Keynote 3 (30 min + 10 min Q&A) • Paper presentations (4 × 10 min + 2 min Q&A) • Closing Remarks (10 min) | Dragi Kocev Bias Variance Labs & Jožef Stefan Institute From Benchmarks to Knowledge Bases: FAIR and Trustworthy Resources for AI4EO | Papers: • ATLAS: Auditing the Geometry of Geophysical Knowledge in EO Foundation Models (Stanislaw Kiedrzynski, Michal Nowicki, Mikolaj Wolinski, Bartlomiej Sobieski, Hubert Baniecki, Przemyslaw Biecek, Agata Wijata) • HELM on BigEarthNet: Benchmarking Hierarchical Multi-Label Land-Cover Classification (Nikola Marić, Marjan Stoimchev, Dragi Kocev, Sašo Džeroski) • VERONA: A framework for for neural network robustness experiments (Aaron Berger, Annelot Bosman, Hendrik Baacke, Holger Hoos, Jan Van Rijn) • A Note on Diversifying Attention for Hierarchical and Explicit Label Modelling with Graph Learning for Multi-Label Image Classification (HELM) (Filip Josifljević, Dubravko Culibrk) |