AI and Earth observation - where to now?
Date:
Talk on physics-based machine learning and its importance for Earth sciences, as part of the collaborative talk on AI and Earth observation - where to now? with Konrad Schindler (ETH, Switzerland), XiaoXiang Zhu (TUM, Germany), Gustau Camps-Valls (Universitat de València, Spain), Mihai Datcu (University POLITEHNICA of Bucharest, Romania), Fabio del Frate (Università di Tor Vergata, Italy), Bertrand Le Saux (DG Connect, European Commission), Devis Tuia (EPFL, Switzerland), Jan van Rijn (Leiden University, the Netherlands) and Nicolas Longépé. at the Living Planet Symposium 2025, taking place on 24 June 2025, in Vienna, Austria.
Abstract: : Artificial Intelligence (AI) has entered all areas of society and Earth Observation (EO) is no exception. The insights contained in multi sensor (multimodal) data can complement and empower traditional physical models and work in unison towards solutions that are accurate, explainable and can enhance scientific discovery. The rise of language models in EO also open new opportunities for interaction with users, or to mine the massive data archive with semantics. In this session, members of the Phi-Lab invited professors will discuss the latest advances and engage critically (and provocatively) with the audience in a forward looking discussion about the future at the interface of AI and remote sensing.