FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386269743> ?p ?o ?g. }

• W4386269743 endingPage "103810" @default.
• W4386269743 startingPage "103810" @default.
• W4386269743 abstract "In this paper, we propose a weakly supervised semantic segmentation method by automatically learning from web images, which are crawled from the Internet by using text queries, without any explicit user annotation or even data filtering. With the goal of handling the massive amount of noisy labels in web images, we design a three-stage approach for weakly-supervised semantic segmentation based on curriculum learning. We first generate pixel-level masks for the training images via a popular weakly-supervised semantic segmentation framework. Then, we consider the noise of the web data in two ways. At the image-level, the complexity of data is measured using its distribution density in a classification feature space. At the pixel-level, the complexity of masks is evaluated by exploring the relationships between saliency map and those segmented images in an unsupervised manner. The key insight to this design is that, common and simple object patterns in images should be salient with both the saliency detector and weakly supervised DCNNs, while they should be sparse with highly regional inconsistency between them. This allows for an efficient implementation of curriculum learning from noisy web images. Experiments on the popular PASCAL VOC 2012 benchmark show that we achieve very competitive performance with scores of 64.0% mIoU using our pure web dataset, which contains noisy, single-label images. We further improve the performance to 69.7% mIoU by using the CurriculumWebSegNet fine-tuned on PASCAL VOC dataset, which has more precise multi-label supervision." @default.
• W4386269743 created "2023-08-31" @default.
• W4386269743 creator A5021559271 @default.
• W4386269743 creator A5039134088 @default.
• W4386269743 creator A5075339911 @default.
• W4386269743 date "2023-11-01" @default.
• W4386269743 modified "2023-10-10" @default.
• W4386269743 title "Webly-supervised semantic segmentation via curriculum learning" @default.
• W4386269743 cites W1495267108 @default.
• W4386269743 cites W1536680647 @default.
• W4386269743 cites W1567302070 @default.
• W4386269743 cites W1927251054 @default.
• W4386269743 cites W2037227137 @default.
• W4386269743 cites W2097117768 @default.
• W4386269743 cites W2102605133 @default.
• W4386269743 cites W2117539524 @default.
• W4386269743 cites W2124219775 @default.
• W4386269743 cites W2133515615 @default.
• W4386269743 cites W2144794286 @default.
• W4386269743 cites W2161185676 @default.
• W4386269743 cites W2165835468 @default.
• W4386269743 cites W2194775991 @default.
• W4386269743 cites W2221898772 @default.
• W4386269743 cites W2294182682 @default.
• W4386269743 cites W2295107390 @default.
• W4386269743 cites W2337429362 @default.
• W4386269743 cites W2412782625 @default.
• W4386269743 cites W2552414813 @default.
• W4386269743 cites W2585747585 @default.
• W4386269743 cites W2600144439 @default.
• W4386269743 cites W2606129492 @default.
• W4386269743 cites W2798683932 @default.
• W4386269743 cites W2799124825 @default.
• W4386269743 cites W2887842788 @default.
• W4386269743 cites W2889081631 @default.
• W4386269743 cites W2901412525 @default.
• W4386269743 cites W2904549000 @default.
• W4386269743 cites W2962758679 @default.
• W4386269743 cites W2963342403 @default.
• W4386269743 cites W2963563573 @default.
• W4386269743 cites W2963876698 @default.
• W4386269743 cites W2981689412 @default.
• W4386269743 cites W2981747647 @default.
• W4386269743 cites W2982093251 @default.
• W4386269743 cites W3010512657 @default.
• W4386269743 cites W3034930876 @default.
• W4386269743 cites W3100040694 @default.
• W4386269743 cites W3104979525 @default.
• W4386269743 cites W3175456851 @default.
• W4386269743 cites W3177958285 @default.
• W4386269743 cites W3183732083 @default.
• W4386269743 cites W3203879378 @default.
• W4386269743 cites W4226158211 @default.
• W4386269743 cites W4239147634 @default.
• W4386269743 cites W4286905173 @default.
• W4386269743 cites W4312566218 @default.
• W4386269743 cites W4312680544 @default.
• W4386269743 cites W4312709262 @default.
• W4386269743 doi "https://doi.org/10.1016/j.cviu.2023.103810" @default.
• W4386269743 hasPublicationYear "2023" @default.
• W4386269743 type Work @default.
• W4386269743 citedByCount "0" @default.
• W4386269743 crossrefType "journal-article" @default.
• W4386269743 hasAuthorship W4386269743A5021559271 @default.
• W4386269743 hasAuthorship W4386269743A5039134088 @default.
• W4386269743 hasAuthorship W4386269743A5075339911 @default.
• W4386269743 hasConcept C119857082 @default.
• W4386269743 hasConcept C13280743 @default.
• W4386269743 hasConcept C136389625 @default.
• W4386269743 hasConcept C153180895 @default.
• W4386269743 hasConcept C154945302 @default.
• W4386269743 hasConcept C185798385 @default.
• W4386269743 hasConcept C199360897 @default.
• W4386269743 hasConcept C205649164 @default.
• W4386269743 hasConcept C2776145971 @default.
• W4386269743 hasConcept C2776321320 @default.
• W4386269743 hasConcept C41008148 @default.
• W4386269743 hasConcept C50644808 @default.
• W4386269743 hasConcept C75608658 @default.
• W4386269743 hasConcept C89600930 @default.
• W4386269743 hasConceptScore W4386269743C119857082 @default.
• W4386269743 hasConceptScore W4386269743C13280743 @default.
• W4386269743 hasConceptScore W4386269743C136389625 @default.
• W4386269743 hasConceptScore W4386269743C153180895 @default.
• W4386269743 hasConceptScore W4386269743C154945302 @default.
• W4386269743 hasConceptScore W4386269743C185798385 @default.
• W4386269743 hasConceptScore W4386269743C199360897 @default.
• W4386269743 hasConceptScore W4386269743C205649164 @default.
• W4386269743 hasConceptScore W4386269743C2776145971 @default.
• W4386269743 hasConceptScore W4386269743C2776321320 @default.
• W4386269743 hasConceptScore W4386269743C41008148 @default.
• W4386269743 hasConceptScore W4386269743C50644808 @default.
• W4386269743 hasConceptScore W4386269743C75608658 @default.
• W4386269743 hasConceptScore W4386269743C89600930 @default.
• W4386269743 hasFunder F4320321001 @default.
• W4386269743 hasFunder F4320335787 @default.
• W4386269743 hasLocation W43862697431 @default.
• W4386269743 hasOpenAccess W4386269743 @default.

• next