KDF reduces catastrophic forgetting in LiDAR place recognition by distilling soft ranking information from the old model and concatenating old and new model features at test time.
Improving point cloud based place recognition with ranking-based loss and large batch training,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Ranking-aware Continual Learning for LiDAR Place Recognition
KDF reduces catastrophic forgetting in LiDAR place recognition by distilling soft ranking information from the old model and concatenating old and new model features at test time.