A training-free monocular pipeline decouples 3D Gaussian geometry from post-hoc object-level open-vocabulary semantics, cutting map memory by about an order of magnitude on Replica.
ARKitSceneRefer: Text- based localization of small objects in diverse real-world 3d indoor scenes
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MonoVoc: Decoupling Geometry and Semantics for Lightweight Monocular Open-Vocabulary 3D Gaussians
A training-free monocular pipeline decouples 3D Gaussian geometry from post-hoc object-level open-vocabulary semantics, cutting map memory by about an order of magnitude on Replica.