Spatial semantic raster matching retrieves the correct map location at 94.5–95.5% Recall@1 under hard synthetic cross-view perturbations, versus 58.6% for a global semantic descriptor.
Visual localization of a UA V using semantic environment information and map data,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SASGeo: Stability-Aware Semantic Map Localization for GNSS-Denied UAVs -- A Framework and Synthetic Proof of Concept
Spatial semantic raster matching retrieves the correct map location at 94.5–95.5% Recall@1 under hard synthetic cross-view perturbations, versus 58.6% for a global semantic descriptor.