Matches in SemOpenAlex for { <https://semopenalex.org/work/W4286233795> ?p ?o ?g. }
Showing items 1 to 75 of
75
with 100 items per page.
- W4286233795 abstract "Since RANSAC, a great deal of research has been devoted to improving both its accuracy and run-time. Still, only a few methods aim at recognizing invalid minimal samples early, before the often expensive model estimation and quality calculation are done. To this end, we propose NeFSAC, an efficient algorithm for neural filtering of motion-inconsistent and poorly-conditioned minimal samples. We train NeFSAC to predict the probability of a minimal sample leading to an accurate relative pose, only based on the pixel coordinates of the image correspondences. Our neural filtering model learns typical motion patterns of samples which lead to unstable poses, and regularities in the possible motions to favour well-conditioned and likely-correct samples. The novel lightweight architecture implements the main invariants of minimal samples for pose estimation, and a novel training scheme addresses the problem of extreme class imbalance. NeFSAC can be plugged into any existing RANSAC-based pipeline. We integrate it into USAC and show that it consistently provides strong speed-ups even under extreme train-test domain gaps - for example, the model trained for the autonomous driving scenario works on PhotoTourism too. We tested NeFSAC on more than 100k image pairs from three publicly available real-world datasets and found that it leads to one order of magnitude speed-up, while often finding more accurate results than USAC alone. The source code is available at https://github.com/cavalli1234/NeFSAC." @default.
- W4286233795 created "2022-07-21" @default.
- W4286233795 creator A5012398971 @default.
- W4286233795 creator A5016636021 @default.
- W4286233795 creator A5021908609 @default.
- W4286233795 date "2022-07-16" @default.
- W4286233795 modified "2023-09-27" @default.
- W4286233795 title "NeFSAC: Neurally Filtered Minimal Samples" @default.
- W4286233795 doi "https://doi.org/10.48550/arxiv.2207.07872" @default.
- W4286233795 hasPublicationYear "2022" @default.
- W4286233795 type Work @default.
- W4286233795 citedByCount "0" @default.
- W4286233795 crossrefType "posted-content" @default.
- W4286233795 hasAuthorship W4286233795A5012398971 @default.
- W4286233795 hasAuthorship W4286233795A5016636021 @default.
- W4286233795 hasAuthorship W4286233795A5021908609 @default.
- W4286233795 hasBestOaLocation W42862337951 @default.
- W4286233795 hasConcept C104114177 @default.
- W4286233795 hasConcept C106131492 @default.
- W4286233795 hasConcept C11413529 @default.
- W4286233795 hasConcept C114744707 @default.
- W4286233795 hasConcept C115961682 @default.
- W4286233795 hasConcept C119857082 @default.
- W4286233795 hasConcept C134306372 @default.
- W4286233795 hasConcept C153180895 @default.
- W4286233795 hasConcept C154945302 @default.
- W4286233795 hasConcept C160633673 @default.
- W4286233795 hasConcept C177264268 @default.
- W4286233795 hasConcept C185592680 @default.
- W4286233795 hasConcept C198531522 @default.
- W4286233795 hasConcept C199360897 @default.
- W4286233795 hasConcept C2776760102 @default.
- W4286233795 hasConcept C31972630 @default.
- W4286233795 hasConcept C33923547 @default.
- W4286233795 hasConcept C36503486 @default.
- W4286233795 hasConcept C41008148 @default.
- W4286233795 hasConcept C43521106 @default.
- W4286233795 hasConcept C43617362 @default.
- W4286233795 hasConceptScore W4286233795C104114177 @default.
- W4286233795 hasConceptScore W4286233795C106131492 @default.
- W4286233795 hasConceptScore W4286233795C11413529 @default.
- W4286233795 hasConceptScore W4286233795C114744707 @default.
- W4286233795 hasConceptScore W4286233795C115961682 @default.
- W4286233795 hasConceptScore W4286233795C119857082 @default.
- W4286233795 hasConceptScore W4286233795C134306372 @default.
- W4286233795 hasConceptScore W4286233795C153180895 @default.
- W4286233795 hasConceptScore W4286233795C154945302 @default.
- W4286233795 hasConceptScore W4286233795C160633673 @default.
- W4286233795 hasConceptScore W4286233795C177264268 @default.
- W4286233795 hasConceptScore W4286233795C185592680 @default.
- W4286233795 hasConceptScore W4286233795C198531522 @default.
- W4286233795 hasConceptScore W4286233795C199360897 @default.
- W4286233795 hasConceptScore W4286233795C2776760102 @default.
- W4286233795 hasConceptScore W4286233795C31972630 @default.
- W4286233795 hasConceptScore W4286233795C33923547 @default.
- W4286233795 hasConceptScore W4286233795C36503486 @default.
- W4286233795 hasConceptScore W4286233795C41008148 @default.
- W4286233795 hasConceptScore W4286233795C43521106 @default.
- W4286233795 hasConceptScore W4286233795C43617362 @default.
- W4286233795 hasLocation W42862337951 @default.
- W4286233795 hasOpenAccess W4286233795 @default.
- W4286233795 hasPrimaryLocation W42862337951 @default.
- W4286233795 hasRelatedWork W10405577 @default.
- W4286233795 hasRelatedWork W12007332 @default.
- W4286233795 hasRelatedWork W12227502 @default.
- W4286233795 hasRelatedWork W12947673 @default.
- W4286233795 hasRelatedWork W2376444 @default.
- W4286233795 hasRelatedWork W3404255 @default.
- W4286233795 hasRelatedWork W5940284 @default.
- W4286233795 hasRelatedWork W6901147 @default.
- W4286233795 hasRelatedWork W7914692 @default.
- W4286233795 hasRelatedWork W8892619 @default.
- W4286233795 isParatext "false" @default.
- W4286233795 isRetracted "false" @default.
- W4286233795 workType "article" @default.