Matches in SemOpenAlex for { <https://semopenalex.org/work/W4205262979> ?p ?o ?g. }
- W4205262979 endingPage "4612" @default.
- W4205262979 startingPage "4599" @default.
- W4205262979 abstract "In this paper, we present a hybrid semantic affinity learning method (HSA) to capture and leverage the dependencies of categories for 3D semantic segmentation. Unlike existing methods that only use the cross-entropy loss to perform one-to-one supervision and ignore the semantic relations between points, our approach aims to learn the label dependencies between 3D points from a hybrid perspective. From a global view, we introduce the structural correlations among different classes to provide global priors for point features. Specifically, we fuse word embeddings of labels and scene-level features as category nodes, which are processed via a graph convolutional network (GCN) to produce the sample-adapted global priors. These priors are then combined with point features to enhance the rationality of semantic predictions. From a local view, we propose the concept of local affinity to effectively model the intra-class and inter-class semantic similarities for adjacent neighborhoods, making the predictions more discriminative. Experimental results show that our method consistently improves the performance of state-of-the-art models across indoor (S3DIS, ScanNet), outdoor (SemanticKITTI), and synthetic (ShapeNet) datasets." @default.
- W4205262979 created "2022-01-26" @default.
- W4205262979 creator A5024102664 @default.
- W4205262979 creator A5028701730 @default.
- W4205262979 creator A5077094563 @default.
- W4205262979 date "2022-07-01" @default.
- W4205262979 modified "2023-10-14" @default.
- W4205262979 title "Learning Hybrid Semantic Affinity for Point Cloud Segmentation" @default.
- W4205262979 cites W1903029394 @default.
- W4205262979 cites W2112467991 @default.
- W4205262979 cites W2143255850 @default.
- W4205262979 cites W2153226019 @default.
- W4205262979 cites W2250539671 @default.
- W4205262979 cites W2412782625 @default.
- W4205262979 cites W2460657278 @default.
- W4205262979 cites W2516803306 @default.
- W4205262979 cites W2532206188 @default.
- W4205262979 cites W2560023338 @default.
- W4205262979 cites W2565662353 @default.
- W4205262979 cites W2594519801 @default.
- W4205262979 cites W2769312834 @default.
- W4205262979 cites W2788158258 @default.
- W4205262979 cites W2795374598 @default.
- W4205262979 cites W2797997528 @default.
- W4205262979 cites W2798380924 @default.
- W4205262979 cites W2888340395 @default.
- W4205262979 cites W2912230690 @default.
- W4205262979 cites W2918473435 @default.
- W4205262979 cites W2932399282 @default.
- W4205262979 cites W2948107928 @default.
- W4205262979 cites W2956542947 @default.
- W4205262979 cites W2960986959 @default.
- W4205262979 cites W2961566087 @default.
- W4205262979 cites W2962676885 @default.
- W4205262979 cites W2962758679 @default.
- W4205262979 cites W2962891704 @default.
- W4205262979 cites W2963125977 @default.
- W4205262979 cites W2963182550 @default.
- W4205262979 cites W2963231572 @default.
- W4205262979 cites W2963281829 @default.
- W4205262979 cites W2963357556 @default.
- W4205262979 cites W2963517242 @default.
- W4205262979 cites W2964051675 @default.
- W4205262979 cites W2968847206 @default.
- W4205262979 cites W2975950300 @default.
- W4205262979 cites W2981983525 @default.
- W4205262979 cites W2983098675 @default.
- W4205262979 cites W2984422960 @default.
- W4205262979 cites W2990479855 @default.
- W4205262979 cites W2990613095 @default.
- W4205262979 cites W2991216808 @default.
- W4205262979 cites W2997216793 @default.
- W4205262979 cites W2997851315 @default.
- W4205262979 cites W3008547073 @default.
- W4205262979 cites W3010797203 @default.
- W4205262979 cites W3012494314 @default.
- W4205262979 cites W3014902535 @default.
- W4205262979 cites W3034239841 @default.
- W4205262979 cites W3034591723 @default.
- W4205262979 cites W3034868495 @default.
- W4205262979 cites W3034978279 @default.
- W4205262979 cites W3035154952 @default.
- W4205262979 cites W3035252826 @default.
- W4205262979 cites W3035339581 @default.
- W4205262979 cites W3035472394 @default.
- W4205262979 cites W3084505175 @default.
- W4205262979 cites W3089874271 @default.
- W4205262979 cites W3105043467 @default.
- W4205262979 cites W3107518100 @default.
- W4205262979 cites W3109944402 @default.
- W4205262979 cites W3110503160 @default.
- W4205262979 cites W3122159272 @default.
- W4205262979 cites W3178663471 @default.
- W4205262979 cites W3184159242 @default.
- W4205262979 doi "https://doi.org/10.1109/tcsvt.2021.3132047" @default.
- W4205262979 hasPublicationYear "2022" @default.
- W4205262979 type Work @default.
- W4205262979 citedByCount "9" @default.
- W4205262979 countsByYear W42052629792022 @default.
- W4205262979 countsByYear W42052629792023 @default.
- W4205262979 crossrefType "journal-article" @default.
- W4205262979 hasAuthorship W4205262979A5024102664 @default.
- W4205262979 hasAuthorship W4205262979A5028701730 @default.
- W4205262979 hasAuthorship W4205262979A5077094563 @default.
- W4205262979 hasConcept C107673813 @default.
- W4205262979 hasConcept C119857082 @default.
- W4205262979 hasConcept C131979681 @default.
- W4205262979 hasConcept C132525143 @default.
- W4205262979 hasConcept C153083717 @default.
- W4205262979 hasConcept C153180895 @default.
- W4205262979 hasConcept C154945302 @default.
- W4205262979 hasConcept C177769412 @default.
- W4205262979 hasConcept C41008148 @default.
- W4205262979 hasConcept C80444323 @default.
- W4205262979 hasConcept C81363708 @default.
- W4205262979 hasConcept C89600930 @default.
- W4205262979 hasConcept C97931131 @default.
- W4205262979 hasConceptScore W4205262979C107673813 @default.