Matches in SemOpenAlex for { <https://semopenalex.org/work/W2788188949> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W2788188949 abstract "We propose a novel locally adaptive learning estimator for enhancing the inter- and intra- discriminative capabilities of Deep Neural Networks, which can be used as improved loss layer for semantic image segmentation tasks. Most loss layers compute pixel-wise cost between feature maps and ground truths, ignoring spatial layouts and interactions between neighboring pixels with same object category, and thus networks cannot be effectively sensitive to intra-class connections. Stride by stride, our method firstly conducts adaptive pooling filter operating over predicted feature maps, aiming to merge predicted distributions over a small group of neighboring pixels with same category, and then it computes cost between the merged distribution vector and their category label. Such design can make groups of neighboring predictions from same category involved into estimations on predicting correctness with respect to their category, and hence train networks to be more sensitive to regional connections between adjacent pixels based on their categories. In the experiments on Pascal VOC 2012 segmentation datasets, the consistently improved results show that our proposed approach achieves better segmentation masks against previous counterparts." @default.
- W2788188949 created "2018-03-06" @default.
- W2788188949 creator A5014442800 @default.
- W2788188949 creator A5054819398 @default.
- W2788188949 creator A5057104359 @default.
- W2788188949 creator A5062689387 @default.
- W2788188949 creator A5085728168 @default.
- W2788188949 date "2018-02-23" @default.
- W2788188949 modified "2023-10-14" @default.
- W2788188949 title "Locally Adaptive Learning Loss for Semantic Image Segmentation" @default.
- W2788188949 cites W1536680647 @default.
- W2788188949 cites W1610060839 @default.
- W2788188949 cites W1686810756 @default.
- W2788188949 cites W2005560667 @default.
- W2788188949 cites W2031489346 @default.
- W2788188949 cites W2037227137 @default.
- W2788188949 cites W2088171019 @default.
- W2788188949 cites W2093717447 @default.
- W2788188949 cites W2138621090 @default.
- W2788188949 cites W2194775991 @default.
- W2788188949 cites W2216125271 @default.
- W2788188949 cites W2271840356 @default.
- W2788188949 cites W2286929393 @default.
- W2788188949 cites W2395611524 @default.
- W2788188949 cites W2412782625 @default.
- W2788188949 cites W2511730936 @default.
- W2788188949 cites W2520774990 @default.
- W2788188949 cites W2607180574 @default.
- W2788188949 cites W2737258237 @default.
- W2788188949 cites W2743473392 @default.
- W2788188949 cites W2762262611 @default.
- W2788188949 cites W2949193792 @default.
- W2788188949 cites W2950053619 @default.
- W2788188949 cites W2953106684 @default.
- W2788188949 cites W2953139137 @default.
- W2788188949 cites W2953236957 @default.
- W2788188949 cites W2963881378 @default.
- W2788188949 cites W3099206234 @default.
- W2788188949 doi "https://doi.org/10.48550/arxiv.1802.08290" @default.
- W2788188949 hasPublicationYear "2018" @default.
- W2788188949 type Work @default.
- W2788188949 sameAs 2788188949 @default.
- W2788188949 citedByCount "2" @default.
- W2788188949 countsByYear W27881889492019 @default.
- W2788188949 countsByYear W27881889492020 @default.
- W2788188949 crossrefType "posted-content" @default.
- W2788188949 hasAuthorship W2788188949A5014442800 @default.
- W2788188949 hasAuthorship W2788188949A5054819398 @default.
- W2788188949 hasAuthorship W2788188949A5057104359 @default.
- W2788188949 hasAuthorship W2788188949A5062689387 @default.
- W2788188949 hasAuthorship W2788188949A5085728168 @default.
- W2788188949 hasBestOaLocation W27881889491 @default.
- W2788188949 hasConcept C11413529 @default.
- W2788188949 hasConcept C138885662 @default.
- W2788188949 hasConcept C153180895 @default.
- W2788188949 hasConcept C154945302 @default.
- W2788188949 hasConcept C160633673 @default.
- W2788188949 hasConcept C199360897 @default.
- W2788188949 hasConcept C2776401178 @default.
- W2788188949 hasConcept C41008148 @default.
- W2788188949 hasConcept C41895202 @default.
- W2788188949 hasConcept C55439883 @default.
- W2788188949 hasConcept C70437156 @default.
- W2788188949 hasConcept C75608658 @default.
- W2788188949 hasConcept C89600930 @default.
- W2788188949 hasConcept C97931131 @default.
- W2788188949 hasConceptScore W2788188949C11413529 @default.
- W2788188949 hasConceptScore W2788188949C138885662 @default.
- W2788188949 hasConceptScore W2788188949C153180895 @default.
- W2788188949 hasConceptScore W2788188949C154945302 @default.
- W2788188949 hasConceptScore W2788188949C160633673 @default.
- W2788188949 hasConceptScore W2788188949C199360897 @default.
- W2788188949 hasConceptScore W2788188949C2776401178 @default.
- W2788188949 hasConceptScore W2788188949C41008148 @default.
- W2788188949 hasConceptScore W2788188949C41895202 @default.
- W2788188949 hasConceptScore W2788188949C55439883 @default.
- W2788188949 hasConceptScore W2788188949C70437156 @default.
- W2788188949 hasConceptScore W2788188949C75608658 @default.
- W2788188949 hasConceptScore W2788188949C89600930 @default.
- W2788188949 hasConceptScore W2788188949C97931131 @default.
- W2788188949 hasLocation W27881889491 @default.
- W2788188949 hasOpenAccess W2788188949 @default.
- W2788188949 hasPrimaryLocation W27881889491 @default.
- W2788188949 hasRelatedWork W1971623867 @default.
- W2788188949 hasRelatedWork W1982770690 @default.
- W2788188949 hasRelatedWork W2124132578 @default.
- W2788188949 hasRelatedWork W2136485282 @default.
- W2788188949 hasRelatedWork W2510758617 @default.
- W2788188949 hasRelatedWork W2546871836 @default.
- W2788188949 hasRelatedWork W2799124825 @default.
- W2788188949 hasRelatedWork W2962136799 @default.
- W2788188949 hasRelatedWork W4281640556 @default.
- W2788188949 hasRelatedWork W2002704236 @default.
- W2788188949 isParatext "false" @default.
- W2788188949 isRetracted "false" @default.
- W2788188949 magId "2788188949" @default.
- W2788188949 workType "article" @default.