A text-guided pruning and completion method on a sparse voxel network achieves state-of-the-art accuracy at 12.43 FPS on ScanRefer, roughly doubling the speed of prior single-stage 3D visual grounding methods.
Scanrefer: 3d object localization in rgb-d scans using natural language
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
dataset 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
dataset 1polarities
use dataset 1representative citing papers
citing papers explorer
-
TSP3D: Text-guided Sparse Voxel Pruning for Efficient 3D Visual Grounding
A text-guided pruning and completion method on a sparse voxel network achieves state-of-the-art accuracy at 12.43 FPS on ScanRefer, roughly doubling the speed of prior single-stage 3D visual grounding methods.