Matches in SemOpenAlex for { <https://semopenalex.org/work/W3109017968> ?p ?o ?g. }
- W3109017968 abstract "The classical matching pipeline used for visual localization typically involves three steps: (i) local feature detection and description, (ii) feature matching, and (iii) outlier rejection. Recently emerged correspondence networks propose to perform those steps inside a single network but suffer from low matching resolution due to the memory bottleneck. In this work, we propose a new perspective to estimate correspondences in a detect-to-refine manner, where we first predict patch-level match proposals and then refine them. We present Patch2Pix, a novel refinement network that refines match proposals by regressing pixel-level matches from the local regions defined by those proposals and jointly rejecting outlier matches with confidence scores. Patch2Pix is weakly supervised to learn correspondences that are consistent with the epipolar geometry of an input image pair. We show that our refinement network significantly improves the performance of correspondence networks on image matching, homography estimation, and localization tasks. In addition, we show that our learned refinement generalizes to fully-supervised methods without re-training, which leads us to state-of-the-art localization performance. The code is available at https://github.com/GrumpyZhou/patch2pix." @default.
- W3109017968 created "2020-12-07" @default.
- W3109017968 creator A5011683384 @default.
- W3109017968 creator A5014776723 @default.
- W3109017968 creator A5090283032 @default.
- W3109017968 date "2020-12-03" @default.
- W3109017968 modified "2023-09-25" @default.
- W3109017968 title "Patch2Pix: Epipolar-Guided Pixel-Level Correspondences" @default.
- W3109017968 cites W1606766208 @default.
- W3109017968 cites W2003447360 @default.
- W3109017968 cites W2033819227 @default.
- W3109017968 cites W2046166954 @default.
- W3109017968 cites W2085261163 @default.
- W3109017968 cites W2086504823 @default.
- W3109017968 cites W2099927035 @default.
- W3109017968 cites W2105303354 @default.
- W3109017968 cites W2108598243 @default.
- W3109017968 cites W2119605622 @default.
- W3109017968 cites W2130017587 @default.
- W3109017968 cites W2158131535 @default.
- W3109017968 cites W2172188317 @default.
- W3109017968 cites W2177274842 @default.
- W3109017968 cites W2471962767 @default.
- W3109017968 cites W2604233003 @default.
- W3109017968 cites W2613718673 @default.
- W3109017968 cites W2740418457 @default.
- W3109017968 cites W2884354140 @default.
- W3109017968 cites W2899771611 @default.
- W3109017968 cites W2962705366 @default.
- W3109017968 cites W2963059198 @default.
- W3109017968 cites W2963157250 @default.
- W3109017968 cites W2963210849 @default.
- W3109017968 cites W2963537932 @default.
- W3109017968 cites W2963674285 @default.
- W3109017968 cites W2963748588 @default.
- W3109017968 cites W2963760790 @default.
- W3109017968 cites W2964073646 @default.
- W3109017968 cites W2964121744 @default.
- W3109017968 cites W2964157791 @default.
- W3109017968 cites W2981799316 @default.
- W3109017968 cites W2987672160 @default.
- W3109017968 cites W2991115635 @default.
- W3109017968 cites W3001017583 @default.
- W3109017968 cites W3034275286 @default.
- W3109017968 cites W3034411221 @default.
- W3109017968 cites W3035563186 @default.
- W3109017968 cites W3092135995 @default.
- W3109017968 cites W3100325658 @default.
- W3109017968 cites W3107540572 @default.
- W3109017968 cites W3210232381 @default.
- W3109017968 doi "https://doi.org/10.48550/arxiv.2012.01909" @default.
- W3109017968 hasPublicationYear "2020" @default.
- W3109017968 type Work @default.
- W3109017968 sameAs 3109017968 @default.
- W3109017968 citedByCount "1" @default.
- W3109017968 countsByYear W31090179682021 @default.
- W3109017968 crossrefType "posted-content" @default.
- W3109017968 hasAuthorship W3109017968A5011683384 @default.
- W3109017968 hasAuthorship W3109017968A5014776723 @default.
- W3109017968 hasAuthorship W3109017968A5090283032 @default.
- W3109017968 hasBestOaLocation W31090179681 @default.
- W3109017968 hasConcept C105795698 @default.
- W3109017968 hasConcept C115961682 @default.
- W3109017968 hasConcept C134306372 @default.
- W3109017968 hasConcept C138885662 @default.
- W3109017968 hasConcept C149635348 @default.
- W3109017968 hasConcept C153180895 @default.
- W3109017968 hasConcept C154945302 @default.
- W3109017968 hasConcept C160633673 @default.
- W3109017968 hasConcept C165064840 @default.
- W3109017968 hasConcept C177264268 @default.
- W3109017968 hasConcept C177846678 @default.
- W3109017968 hasConcept C199360897 @default.
- W3109017968 hasConcept C23379248 @default.
- W3109017968 hasConcept C2776401178 @default.
- W3109017968 hasConcept C2776760102 @default.
- W3109017968 hasConcept C2780427248 @default.
- W3109017968 hasConcept C2780513914 @default.
- W3109017968 hasConcept C28751775 @default.
- W3109017968 hasConcept C31972630 @default.
- W3109017968 hasConcept C33923547 @default.
- W3109017968 hasConcept C41008148 @default.
- W3109017968 hasConcept C41895202 @default.
- W3109017968 hasConcept C43521106 @default.
- W3109017968 hasConcept C75280867 @default.
- W3109017968 hasConcept C79337645 @default.
- W3109017968 hasConceptScore W3109017968C105795698 @default.
- W3109017968 hasConceptScore W3109017968C115961682 @default.
- W3109017968 hasConceptScore W3109017968C134306372 @default.
- W3109017968 hasConceptScore W3109017968C138885662 @default.
- W3109017968 hasConceptScore W3109017968C149635348 @default.
- W3109017968 hasConceptScore W3109017968C153180895 @default.
- W3109017968 hasConceptScore W3109017968C154945302 @default.
- W3109017968 hasConceptScore W3109017968C160633673 @default.
- W3109017968 hasConceptScore W3109017968C165064840 @default.
- W3109017968 hasConceptScore W3109017968C177264268 @default.
- W3109017968 hasConceptScore W3109017968C177846678 @default.
- W3109017968 hasConceptScore W3109017968C199360897 @default.
- W3109017968 hasConceptScore W3109017968C23379248 @default.
- W3109017968 hasConceptScore W3109017968C2776401178 @default.