Dense matches are summarized into sparse representative matches with 9x9 constraint matrices, giving comparable pose accuracy at 10-100x lower RANSAC runtime.
MAGSAC: marginalizing sample consensus
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
-
Dense Match Summarization for Faster Two-view Estimation
Dense matches are summarized into sparse representative matches with 9x9 constraint matrices, giving comparable pose accuracy at 10-100x lower RANSAC runtime.