GeoMEB

New Geo-Embed model unifies multimodal urban understanding tasks

Researchers have introduced Geo-Embed, a novel multimodal embedding model designed for urban understanding tasks. This model adapts a shared vision-language backbone to handle diverse geospatial inputs, including images…