Where on Earth was an image taken—and how do we align ground-level views with overhead reference data? Image localization is a long-standing focus of the lab, spanning static outdoor cameras, single-image geolocalization, and cross-view matching between street-level and satellite imagery. Recent work includes goal modality agnostic active geo-localization (GOMAA-Geo), structure-aware direct pose estimation, and localizing limited field-of-view images via cross-view matching (ArcGeo). These methods connect ground photos, panoramas, and maps into a shared geospatial frame for search, monitoring, and embodied exploration.
Spotlight publications tagged for this research area.