FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3202299736> ?p ?o ?g. }

• W3202299736 abstract "Weakly supervised semantic segmentation (WSSS) using image-level classification labels usually utilizes the Class Activation Maps (CAMs) to localize objects of interest in images. While pointing out that CAMs only highlight the most discriminative regions of the classes of interest, adversarial erasing (AE) methods have been proposed to further explore the less discriminative regions. In this paper, we review the potential of the pre-trained classifier which is trained on the raw images. We experimentally verify that the ordinary classifier <sup xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>1</sup> already has the capability to activate the less discriminative regions if the most discriminative regions are erased to some extent. Based on that, we propose a class-specific AE-based framework that fully exploits the potential of an ordinary classifier. Our framework (1) adopts the ordinary classifier to notify the regions to be erased and (2) generates a class-specific mask for erasing by randomly sampling a single specific class to be erased (target class) among the existing classes on the image for obtaining more precise CAMs. Specifically, with the guidance of the ordinary classifier, the proposed CAMs Generation Network (CGNet) is enforced to generate a CAM of the target class while constraining the CAM not to intrude the object regions of the other classes. Along with the pseudo-labels refined from our CAMs, we achieve the state-of-the-art WSSS performance on both PASCAL VOC 2012 and MS-COCO dataset only with image-level supervision. The code is available at https://github.com/KAIST-vilab/OC-CSE." @default.
• W3202299736 created "2021-10-11" @default.
• W3202299736 creator A5012837410 @default.
• W3202299736 creator A5042808616 @default.
• W3202299736 creator A5059067295 @default.
• W3202299736 creator A5066751134 @default.
• W3202299736 creator A5084746226 @default.
• W3202299736 date "2021-10-01" @default.
• W3202299736 modified "2023-10-03" @default.
• W3202299736 title "Unlocking the Potential of Ordinary Classifier: Class-specific Adversarial Erasing Framework for Weakly Supervised Semantic Segmentation" @default.
• W3202299736 cites W1495267108 @default.
• W3202299736 cites W2031489346 @default.
• W3202299736 cites W2117539524 @default.
• W3202299736 cites W2221898772 @default.
• W3202299736 cites W2295107390 @default.
• W3202299736 cites W2412782625 @default.
• W3202299736 cites W2552414813 @default.
• W3202299736 cites W2558580397 @default.
• W3202299736 cites W2600144439 @default.
• W3202299736 cites W2618530766 @default.
• W3202299736 cites W2798376494 @default.
• W3202299736 cites W2798715809 @default.
• W3202299736 cites W2799124825 @default.
• W3202299736 cites W2890498246 @default.
• W3202299736 cites W2962758679 @default.
• W3202299736 cites W2962867364 @default.
• W3202299736 cites W2963670239 @default.
• W3202299736 cites W2964254867 @default.
• W3202299736 cites W2964274719 @default.
• W3202299736 cites W2980189057 @default.
• W3202299736 cites W2982093251 @default.
• W3202299736 cites W2991083560 @default.
• W3202299736 cites W2996952120 @default.
• W3202299736 cites W2997851315 @default.
• W3202299736 cites W3034333089 @default.
• W3202299736 cites W3034373787 @default.
• W3202299736 cites W3034930876 @default.
• W3202299736 doi "https://doi.org/10.1109/iccv48922.2021.00691" @default.
• W3202299736 hasPublicationYear "2021" @default.
• W3202299736 type Work @default.
• W3202299736 sameAs 3202299736 @default.
• W3202299736 citedByCount "29" @default.
• W3202299736 countsByYear W32022997362021 @default.
• W3202299736 countsByYear W32022997362022 @default.
• W3202299736 countsByYear W32022997362023 @default.
• W3202299736 crossrefType "proceedings-article" @default.
• W3202299736 hasAuthorship W3202299736A5012837410 @default.
• W3202299736 hasAuthorship W3202299736A5042808616 @default.
• W3202299736 hasAuthorship W3202299736A5059067295 @default.
• W3202299736 hasAuthorship W3202299736A5066751134 @default.
• W3202299736 hasAuthorship W3202299736A5084746226 @default.
• W3202299736 hasConcept C115961682 @default.
• W3202299736 hasConcept C119857082 @default.
• W3202299736 hasConcept C153180895 @default.
• W3202299736 hasConcept C154945302 @default.
• W3202299736 hasConcept C37736160 @default.
• W3202299736 hasConcept C41008148 @default.
• W3202299736 hasConcept C75294576 @default.
• W3202299736 hasConcept C89600930 @default.
• W3202299736 hasConcept C95623464 @default.
• W3202299736 hasConcept C97931131 @default.
• W3202299736 hasConceptScore W3202299736C115961682 @default.
• W3202299736 hasConceptScore W3202299736C119857082 @default.
• W3202299736 hasConceptScore W3202299736C153180895 @default.
• W3202299736 hasConceptScore W3202299736C154945302 @default.
• W3202299736 hasConceptScore W3202299736C37736160 @default.
• W3202299736 hasConceptScore W3202299736C41008148 @default.
• W3202299736 hasConceptScore W3202299736C75294576 @default.
• W3202299736 hasConceptScore W3202299736C89600930 @default.
• W3202299736 hasConceptScore W3202299736C95623464 @default.
• W3202299736 hasConceptScore W3202299736C97931131 @default.
• W3202299736 hasLocation W32022997361 @default.
• W3202299736 hasOpenAccess W3202299736 @default.
• W3202299736 hasPrimaryLocation W32022997361 @default.
• W3202299736 hasRelatedWork W2024160000 @default.
• W3202299736 hasRelatedWork W2061273563 @default.
• W3202299736 hasRelatedWork W2129933262 @default.
• W3202299736 hasRelatedWork W2285052147 @default.
• W3202299736 hasRelatedWork W2563096758 @default.
• W3202299736 hasRelatedWork W2729514902 @default.
• W3202299736 hasRelatedWork W2742991909 @default.
• W3202299736 hasRelatedWork W2773500201 @default.
• W3202299736 hasRelatedWork W2905846897 @default.
• W3202299736 hasRelatedWork W2972035100 @default.
• W3202299736 isParatext "false" @default.
• W3202299736 isRetracted "false" @default.
• W3202299736 magId "3202299736" @default.
• W3202299736 workType "article" @default.