A satellite-free training framework reconstructs 3D drone scenes via Gaussian splatting, generates geometry-normalized pseudo-orthophotos, and aggregates DINOv3 features with a Fisher vector model trained only on drone data to enable cross-view retrieval.
3dgs lsr: Large scale relocation for autonomous driving based on 3d gaussian splatting
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 2years
2026 2roles
background 1polarities
background 1representative citing papers
LSGS-Loc delivers state-of-the-art accuracy and robustness for 3DGS-based visual localization in large UAV scenes via scale-aware initialization and reliability masking without scene-specific training.
citing papers explorer
-
Satellite-Free Training for Drone-View Geo-Localization
A satellite-free training framework reconstructs 3D drone scenes via Gaussian splatting, generates geometry-normalized pseudo-orthophotos, and aggregates DINOv3 features with a Fisher vector model trained only on drone data to enable cross-view retrieval.
-
LSGS-Loc: Towards Robust 3DGS-Based Visual Localization for Large-Scale UAV Scenarios
LSGS-Loc delivers state-of-the-art accuracy and robustness for 3DGS-based visual localization in large UAV scenes via scale-aware initialization and reliability masking without scene-specific training.