Matches in SemOpenAlex for { <https://semopenalex.org/work/W4226337479> ?p ?o ?g. }
- W4226337479 endingPage "2105" @default.
- W4226337479 startingPage "2094" @default.
- W4226337479 abstract "The goal of ground-to-aerial image geo-localization is to determine the location of a ground query image by matching it against a reference database consisting of aerial/satellite images. This task is highly challenging due to the large appearance difference caused by extreme changes in viewpoint and orientation. In this work, we show that the training difficulty is an important cue that can be leveraged to improve metric learning on cross-view images. More specifically, we propose a new Soft Exemplar Highlighting (SEH) loss to achieve online soft selection of exemplars. Adaptive weights are generated for exemplars by measuring their associated training difficulty using distance rectified logistic regression. These weights are then constrained to remove simple exemplars from training and truncate the large weights of extremely hard exemplars to escape from the trap with a local optimal solution. We further use the proposed SEH loss to train two mainstream convolutional neural networks for ground-to-aerial image-based geo-localization. Experimental results on two benchmark cross-view image datasets demonstrate that the proposed method achieves significant improvements in feature discriminativeness and outperforms the state-of-the-art image-based geo-localization methods." @default.
- W4226337479 created "2022-05-05" @default.
- W4226337479 creator A5009750573 @default.
- W4226337479 creator A5032533885 @default.
- W4226337479 creator A5047532570 @default.
- W4226337479 creator A5068710717 @default.
- W4226337479 creator A5075023879 @default.
- W4226337479 date "2022-01-01" @default.
- W4226337479 modified "2023-10-18" @default.
- W4226337479 title "Soft Exemplar Highlighting for Cross-View Image-Based Geo-Localization" @default.
- W4226337479 cites W1162411702 @default.
- W4226337479 cites W1552672384 @default.
- W4226337479 cites W1946093182 @default.
- W4226337479 cites W1981154748 @default.
- W4226337479 cites W1988873224 @default.
- W4226337479 cites W1998728242 @default.
- W4226337479 cites W2061153075 @default.
- W4226337479 cites W2081418428 @default.
- W4226337479 cites W2117228865 @default.
- W4226337479 cites W2151103935 @default.
- W4226337479 cites W2199890863 @default.
- W4226337479 cites W2217115543 @default.
- W4226337479 cites W2340296683 @default.
- W4226337479 cites W2468801144 @default.
- W4226337479 cites W2471048925 @default.
- W4226337479 cites W2479919622 @default.
- W4226337479 cites W2526833393 @default.
- W4226337479 cites W2545896075 @default.
- W4226337479 cites W2572697301 @default.
- W4226337479 cites W2598199894 @default.
- W4226337479 cites W2620629206 @default.
- W4226337479 cites W2737075200 @default.
- W4226337479 cites W2738649458 @default.
- W4226337479 cites W2767743348 @default.
- W4226337479 cites W2799087793 @default.
- W4226337479 cites W2886324878 @default.
- W4226337479 cites W2887134078 @default.
- W4226337479 cites W2951019013 @default.
- W4226337479 cites W2961917444 @default.
- W4226337479 cites W2963474852 @default.
- W4226337479 cites W2981670313 @default.
- W4226337479 cites W2992077842 @default.
- W4226337479 cites W2997129483 @default.
- W4226337479 cites W3035158519 @default.
- W4226337479 cites W3047057232 @default.
- W4226337479 cites W3081227581 @default.
- W4226337479 cites W3099206234 @default.
- W4226337479 cites W3111205756 @default.
- W4226337479 cites W3128990302 @default.
- W4226337479 cites W3174800995 @default.
- W4226337479 cites W4205474609 @default.
- W4226337479 cites W639708223 @default.
- W4226337479 doi "https://doi.org/10.1109/tip.2022.3152046" @default.
- W4226337479 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/35196234" @default.
- W4226337479 hasPublicationYear "2022" @default.
- W4226337479 type Work @default.
- W4226337479 citedByCount "11" @default.
- W4226337479 countsByYear W42263374792022 @default.
- W4226337479 countsByYear W42263374792023 @default.
- W4226337479 crossrefType "journal-article" @default.
- W4226337479 hasAuthorship W4226337479A5009750573 @default.
- W4226337479 hasAuthorship W4226337479A5032533885 @default.
- W4226337479 hasAuthorship W4226337479A5047532570 @default.
- W4226337479 hasAuthorship W4226337479A5068710717 @default.
- W4226337479 hasAuthorship W4226337479A5075023879 @default.
- W4226337479 hasConcept C105795698 @default.
- W4226337479 hasConcept C115961682 @default.
- W4226337479 hasConcept C13280743 @default.
- W4226337479 hasConcept C153180895 @default.
- W4226337479 hasConcept C154945302 @default.
- W4226337479 hasConcept C162324750 @default.
- W4226337479 hasConcept C16345878 @default.
- W4226337479 hasConcept C165064840 @default.
- W4226337479 hasConcept C176217482 @default.
- W4226337479 hasConcept C185798385 @default.
- W4226337479 hasConcept C205649164 @default.
- W4226337479 hasConcept C21547014 @default.
- W4226337479 hasConcept C2524010 @default.
- W4226337479 hasConcept C31972630 @default.
- W4226337479 hasConcept C33923547 @default.
- W4226337479 hasConcept C41008148 @default.
- W4226337479 hasConcept C52622490 @default.
- W4226337479 hasConcept C81363708 @default.
- W4226337479 hasConceptScore W4226337479C105795698 @default.
- W4226337479 hasConceptScore W4226337479C115961682 @default.
- W4226337479 hasConceptScore W4226337479C13280743 @default.
- W4226337479 hasConceptScore W4226337479C153180895 @default.
- W4226337479 hasConceptScore W4226337479C154945302 @default.
- W4226337479 hasConceptScore W4226337479C162324750 @default.
- W4226337479 hasConceptScore W4226337479C16345878 @default.
- W4226337479 hasConceptScore W4226337479C165064840 @default.
- W4226337479 hasConceptScore W4226337479C176217482 @default.
- W4226337479 hasConceptScore W4226337479C185798385 @default.
- W4226337479 hasConceptScore W4226337479C205649164 @default.
- W4226337479 hasConceptScore W4226337479C21547014 @default.
- W4226337479 hasConceptScore W4226337479C2524010 @default.
- W4226337479 hasConceptScore W4226337479C31972630 @default.
- W4226337479 hasConceptScore W4226337479C33923547 @default.