Matches in SemOpenAlex for { <https://semopenalex.org/work/W2770660904> ?p ?o ?g. }
- W2770660904 abstract "We present an accurate method for estimation of the affine shape of local features. The method is trained in a novel way, exploiting the recently proposed HardNet triplet loss. The loss function is driven by patch descriptor differences, avoiding problems with symmetries. Moreover, such training process does not require precisely geometrically aligned patches. The affine shape is represented in a way amenable to learning by stochastic gradient descent. When plugged into a state-of-the-art wide baseline matching algorithm, the performance on standard datasets improves in both the number of challenging pairs matched and the number of inliers. Finally, AffNet with combination of Hessian detector and HardNet descriptor improves bag-of-visual-words based state of the art on Oxford5k and Paris6k by large margin, 4.5 and 4.2 mAP points respectively. The source code and trained networks are available at this https URL" @default.
- W2770660904 created "2017-12-04" @default.
- W2770660904 creator A5007656938 @default.
- W2770660904 creator A5035547460 @default.
- W2770660904 creator A5038939197 @default.
- W2770660904 date "2017-11-17" @default.
- W2770660904 modified "2023-10-17" @default.
- W2770660904 title "Learning Discriminative Affine Regions via Discriminability." @default.
- W2770660904 cites W1534408538 @default.
- W2770660904 cites W1606766208 @default.
- W2770660904 cites W1627400044 @default.
- W2770660904 cites W1646493568 @default.
- W2770660904 cites W1908016767 @default.
- W2770660904 cites W1929856797 @default.
- W2770660904 cites W1945298332 @default.
- W2770660904 cites W1972378554 @default.
- W2770660904 cites W1979931042 @default.
- W2770660904 cites W2068143350 @default.
- W2770660904 cites W2083833836 @default.
- W2770660904 cites W2088866137 @default.
- W2770660904 cites W2095705004 @default.
- W2770660904 cites W2099363089 @default.
- W2770660904 cites W2118877769 @default.
- W2770660904 cites W2126060993 @default.
- W2770660904 cites W2131846894 @default.
- W2770660904 cites W2141362318 @default.
- W2770660904 cites W2144796873 @default.
- W2770660904 cites W2148809531 @default.
- W2770660904 cites W2151103935 @default.
- W2770660904 cites W2156365280 @default.
- W2770660904 cites W2160341214 @default.
- W2770660904 cites W2160614266 @default.
- W2770660904 cites W2165497495 @default.
- W2770660904 cites W2418033038 @default.
- W2770660904 cites W2426541698 @default.
- W2770660904 cites W2471962767 @default.
- W2770660904 cites W2544587078 @default.
- W2770660904 cites W2559091987 @default.
- W2770660904 cites W2586073474 @default.
- W2770660904 cites W2608303904 @default.
- W2770660904 cites W2612112834 @default.
- W2770660904 cites W2614218061 @default.
- W2770660904 cites W2618039218 @default.
- W2770660904 cites W2737094507 @default.
- W2770660904 cites W2737260104 @default.
- W2770660904 cites W2749224737 @default.
- W2770660904 cites W2767621619 @default.
- W2770660904 cites W2772747295 @default.
- W2770660904 cites W2795296309 @default.
- W2770660904 cites W2899771611 @default.
- W2770660904 cites W2911652069 @default.
- W2770660904 cites W2949117887 @default.
- W2770660904 cites W2949213045 @default.
- W2770660904 cites W2949650786 @default.
- W2770660904 cites W2949896259 @default.
- W2770660904 cites W2950974964 @default.
- W2770660904 cites W2951628298 @default.
- W2770660904 cites W2952052436 @default.
- W2770660904 cites W2964121744 @default.
- W2770660904 hasPublicationYear "2017" @default.
- W2770660904 type Work @default.
- W2770660904 sameAs 2770660904 @default.
- W2770660904 citedByCount "2" @default.
- W2770660904 countsByYear W27706609042019 @default.
- W2770660904 crossrefType "posted-content" @default.
- W2770660904 hasAuthorship W2770660904A5007656938 @default.
- W2770660904 hasAuthorship W2770660904A5035547460 @default.
- W2770660904 hasAuthorship W2770660904A5038939197 @default.
- W2770660904 hasConcept C105795698 @default.
- W2770660904 hasConcept C11413529 @default.
- W2770660904 hasConcept C119857082 @default.
- W2770660904 hasConcept C153180895 @default.
- W2770660904 hasConcept C154945302 @default.
- W2770660904 hasConcept C165064840 @default.
- W2770660904 hasConcept C177264268 @default.
- W2770660904 hasConcept C199360897 @default.
- W2770660904 hasConcept C203616005 @default.
- W2770660904 hasConcept C2524010 @default.
- W2770660904 hasConcept C2776760102 @default.
- W2770660904 hasConcept C28826006 @default.
- W2770660904 hasConcept C33923547 @default.
- W2770660904 hasConcept C41008148 @default.
- W2770660904 hasConcept C774472 @default.
- W2770660904 hasConcept C92757383 @default.
- W2770660904 hasConcept C97931131 @default.
- W2770660904 hasConceptScore W2770660904C105795698 @default.
- W2770660904 hasConceptScore W2770660904C11413529 @default.
- W2770660904 hasConceptScore W2770660904C119857082 @default.
- W2770660904 hasConceptScore W2770660904C153180895 @default.
- W2770660904 hasConceptScore W2770660904C154945302 @default.
- W2770660904 hasConceptScore W2770660904C165064840 @default.
- W2770660904 hasConceptScore W2770660904C177264268 @default.
- W2770660904 hasConceptScore W2770660904C199360897 @default.
- W2770660904 hasConceptScore W2770660904C203616005 @default.
- W2770660904 hasConceptScore W2770660904C2524010 @default.
- W2770660904 hasConceptScore W2770660904C2776760102 @default.
- W2770660904 hasConceptScore W2770660904C28826006 @default.
- W2770660904 hasConceptScore W2770660904C33923547 @default.
- W2770660904 hasConceptScore W2770660904C41008148 @default.
- W2770660904 hasConceptScore W2770660904C774472 @default.