Matches in SemOpenAlex for { <https://semopenalex.org/work/W2901048985> ?p ?o ?g. }
- W2901048985 abstract "Building a large image dataset with high-quality object masks for semantic segmentation is costly and time consuming. In this paper, we introduce a principled semi-supervised framework that only uses a small set of fully supervised images (having semantic segmentation labels and box labels) and a set of images with only object bounding box labels (we call it the weak set). Our framework trains the primary segmentation model with the aid of an ancillary model that generates initial segmentation labels for the weak set and a self-correction module that improves the generated labels during training using the increasingly accurate primary model. We introduce two variants of the self-correction module using either linear or convolutional functions. Experiments on the PASCAL VOC 2012 and Cityscape datasets show that our models trained with a small fully supervised set perform similar to, or better than, models trained with a large fully supervised set while requiring ~7x less annotation effort." @default.
- W2901048985 created "2018-11-29" @default.
- W2901048985 creator A5036504011 @default.
- W2901048985 creator A5038984764 @default.
- W2901048985 creator A5076214952 @default.
- W2901048985 date "2018-11-17" @default.
- W2901048985 modified "2023-09-27" @default.
- W2901048985 title "Weakly Supervised Semantic Image Segmentation with Self-correcting Networks" @default.
- W2901048985 cites W1495267108 @default.
- W2901048985 cites W1522734439 @default.
- W2901048985 cites W1817277359 @default.
- W2901048985 cites W1866072925 @default.
- W2901048985 cites W1896424170 @default.
- W2901048985 cites W1901129140 @default.
- W2901048985 cites W1901229278 @default.
- W2901048985 cites W1903029394 @default.
- W2901048985 cites W1910657905 @default.
- W2901048985 cites W1921293667 @default.
- W2901048985 cites W1927251054 @default.
- W2901048985 cites W1947481528 @default.
- W2901048985 cites W1976948919 @default.
- W2901048985 cites W2016053056 @default.
- W2901048985 cites W2037227137 @default.
- W2901048985 cites W2102492119 @default.
- W2901048985 cites W2113005879 @default.
- W2901048985 cites W2113290770 @default.
- W2901048985 cites W2124351162 @default.
- W2901048985 cites W2124592697 @default.
- W2901048985 cites W2144794286 @default.
- W2901048985 cites W2156303437 @default.
- W2901048985 cites W2161236525 @default.
- W2901048985 cites W2162915993 @default.
- W2901048985 cites W2163605009 @default.
- W2901048985 cites W2194775991 @default.
- W2901048985 cites W2221898772 @default.
- W2901048985 cites W2252268321 @default.
- W2901048985 cites W2308045930 @default.
- W2901048985 cites W2337429362 @default.
- W2901048985 cites W2340897893 @default.
- W2901048985 cites W2412782625 @default.
- W2901048985 cites W2531409750 @default.
- W2901048985 cites W2552414813 @default.
- W2901048985 cites W2560023338 @default.
- W2901048985 cites W2563705555 @default.
- W2901048985 cites W2577784528 @default.
- W2901048985 cites W2585747585 @default.
- W2901048985 cites W2600144439 @default.
- W2901048985 cites W2613718673 @default.
- W2901048985 cites W2630837129 @default.
- W2901048985 cites W2738853914 @default.
- W2901048985 cites W2775306753 @default.
- W2901048985 cites W2778764040 @default.
- W2901048985 cites W2798376494 @default.
- W2901048985 cites W2799124825 @default.
- W2901048985 cites W2885139206 @default.
- W2901048985 cites W2951183276 @default.
- W2901048985 cites W2952289000 @default.
- W2901048985 cites W2956648669 @default.
- W2901048985 cites W2962749380 @default.
- W2901048985 cites W2962758679 @default.
- W2901048985 cites W2963037989 @default.
- W2901048985 cites W2963108253 @default.
- W2901048985 cites W2963311325 @default.
- W2901048985 cites W2963371670 @default.
- W2901048985 cites W2963386594 @default.
- W2901048985 cites W2963483401 @default.
- W2901048985 cites W2963563573 @default.
- W2901048985 cites W2963697527 @default.
- W2901048985 cites W2963759070 @default.
- W2901048985 cites W2963798052 @default.
- W2901048985 cites W2963840672 @default.
- W2901048985 cites W2963921921 @default.
- W2901048985 cites W2964014798 @default.
- W2901048985 cites W2964062189 @default.
- W2901048985 cites W2964081807 @default.
- W2901048985 cites W2964096266 @default.
- W2901048985 cites W2964274690 @default.
- W2901048985 cites W2964288706 @default.
- W2901048985 cites W2964292098 @default.
- W2901048985 cites W2964309882 @default.
- W2901048985 cites W2980189057 @default.
- W2901048985 cites W2982631194 @default.
- W2901048985 cites W298976848 @default.
- W2901048985 cites W3106250896 @default.
- W2901048985 cites W611457968 @default.
- W2901048985 hasPublicationYear "2018" @default.
- W2901048985 type Work @default.
- W2901048985 sameAs 2901048985 @default.
- W2901048985 citedByCount "1" @default.
- W2901048985 countsByYear W29010489852021 @default.
- W2901048985 crossrefType "posted-content" @default.
- W2901048985 hasAuthorship W2901048985A5036504011 @default.
- W2901048985 hasAuthorship W2901048985A5038984764 @default.
- W2901048985 hasAuthorship W2901048985A5076214952 @default.
- W2901048985 hasConcept C115961682 @default.
- W2901048985 hasConcept C119857082 @default.
- W2901048985 hasConcept C124504099 @default.
- W2901048985 hasConcept C147037132 @default.
- W2901048985 hasConcept C153180895 @default.
- W2901048985 hasConcept C154945302 @default.