MVL-Loc fuses CLIP image and text features with per-scene pose heads and reports state-of-the-art multi-scene relocalization accuracy on 7Scenes and Cambridge Landmarks.
Atloc: Attention guided camera localization,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
MVL-Loc: Leveraging Vision-Language Model for Generalizable Multi-Scene Camera Relocalization
MVL-Loc fuses CLIP image and text features with per-scene pose heads and reports state-of-the-art multi-scene relocalization accuracy on 7Scenes and Cambridge Landmarks.