10-01, 14:00–14:30 (Europe/London), Create 1
This talk will present our progress in spatial representation learning across several widely used geospatial data sources, gradually leading to our recent work on developing geospatial foundation models, particularly models pretrained using OpenStreetMap.
The rise of large-scale pre-trained models, also known as foundation models, has sparked great interest within the geospatial community, to develop foundation models that can be used for various geospatial analytical tasks. To this end, we view learning effective representations from multi-modal geospatial data as the cornerstone of developing geospatial foundation models. This talk will then present our progress in spatial representation learning across several widely used geospatial data sources, gradually leading to our recent work on developing geospatial foundation models, particularly models pretrained using OpenStreetMap.
Weiming Huang is a Lecturer at the School of Geography, University of Leeds. He obtained his PhD at Lund University, Sweden. With a mixed background in GIScience and Computer Science, his research interests include spatial data mining, geospatial foundation models, and knowledge graphs.