DeepMind and WRI publish ForTy, a global 2020 forest-type map
ForTy labels six land classes at 10-meter resolution and was judged correct on 90.2 percent of 8,190 sites.
2
Stories per day · 30 days
1 RSS
Relatedcontrastive-learning2location-encoders2multimodal2eudr1forest-mapping1google-deepmind1satellite1sentinel1sentinel-11sentinel-21
6 stories, newest first · one card per story
Stories tagged #geospatial
LoCA aligns location encoders with multimodal satellite imagery
On 17 tasks LoCA has the most wins and is the only model ahead on a majority of tasks.
LoCA co-aligns Sentinel encoders for geospatial location encoding
On 17 tasks LoCA wins 68.8% of per-task comparisons and is the only encoder ahead of every rival on a majority of tasks.
Mercator ↔ Equal Earth
Simon Willison used GPT-6 Astra (medium) in ChatGPT Work to build an animated D3 transition between Mercator and Equal Earth map projections.
GeoNetwork Fixes Unauthenticated RCE Chain Affecting Government Geoportal Backends
GeoNetwork patched two chained flaws (CVE-2026-63219, CVE-2026-58400) enabling unauthenticated RCE in government geospatial metadata catalogs.
1
Introducing OlmoEarth embeddings: Custom embedding exports from OlmoEarth Studio for downstream analysis
AllenAI's OlmoEarth Studio adds custom embedding exports to support downstream geospatial analysis workflows.