QCSA reduces inserted loop factors 3.8 times and raises precision from 0.542 to 0.717 on the SNULib dataset while lowering worst-case trajectory error compared with dense Top-1+G-ICP baselines.
Pointnetvlad: Deep point cloud based retrieval for large-scale place recognition
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Query-Calibrated Segmental Admission for Descriptor-Agnostic LiDAR Loop Closure in Repetitive Environments
QCSA reduces inserted loop factors 3.8 times and raises precision from 0.542 to 0.717 on the SNULib dataset while lowering worst-case trajectory error compared with dense Top-1+G-ICP baselines.