Matches in SemOpenAlex for { <https://semopenalex.org/work/W3202397162> ?p ?o ?g. }
- W3202397162 abstract "Domain generalization for semantic segmentation is highly demanded in real applications, where a trained model is expected to work well in previously unseen domains. One challenge lies in the lack of data which could cover the diverse distributions of the possible unseen domains for training. In this paper, we propose a WEb-image assisted Domain GEneralization (WEDGE) scheme, which is the first to exploit the diversity of web-crawled images for generalizable semantic segmentation. To explore and exploit the real-world data distributions, we collect web-crawled images which present large diversity in terms of weather conditions, sites, lighting, camera styles, etc. We also present a method which injects styles of the web-crawled images into training images on-the-fly during training, which enables the network to experience images of diverse styles with reliable labels for effective training. Moreover, we use the web-crawled images with their predicted pseudo labels for training to further enhance the capability of the network. Extensive experiments demonstrate that our method clearly outperforms existing domain generalization techniques." @default.
- W3202397162 created "2021-10-11" @default.
- W3202397162 creator A5009948118 @default.
- W3202397162 creator A5049963367 @default.
- W3202397162 creator A5051469345 @default.
- W3202397162 creator A5060343759 @default.
- W3202397162 creator A5074668738 @default.
- W3202397162 date "2021-09-29" @default.
- W3202397162 modified "2023-10-16" @default.
- W3202397162 title "WEDGE: Web-Image Assisted Domain Generalization for Semantic Segmentation" @default.
- W3202397162 cites W1964763677 @default.
- W3202397162 cites W1973054923 @default.
- W3202397162 cites W2081613070 @default.
- W3202397162 cites W2108598243 @default.
- W3202397162 cites W2124219775 @default.
- W3202397162 cites W2155858138 @default.
- W3202397162 cites W2194775991 @default.
- W3202397162 cites W2331128040 @default.
- W3202397162 cites W2340897893 @default.
- W3202397162 cites W2346746376 @default.
- W3202397162 cites W2412782625 @default.
- W3202397162 cites W2431874326 @default.
- W3202397162 cites W2475287302 @default.
- W3202397162 cites W2487365028 @default.
- W3202397162 cites W2562192638 @default.
- W3202397162 cites W2603777577 @default.
- W3202397162 cites W2739759330 @default.
- W3202397162 cites W2763549966 @default.
- W3202397162 cites W2781228439 @default.
- W3202397162 cites W2798658180 @default.
- W3202397162 cites W2798913983 @default.
- W3202397162 cites W2884366600 @default.
- W3202397162 cites W2895281799 @default.
- W3202397162 cites W2962793481 @default.
- W3202397162 cites W2962835968 @default.
- W3202397162 cites W2962839335 @default.
- W3202397162 cites W2963043696 @default.
- W3202397162 cites W2963073217 @default.
- W3202397162 cites W2963107255 @default.
- W3202397162 cites W2969893028 @default.
- W3202397162 cites W2972285644 @default.
- W3202397162 cites W2981429991 @default.
- W3202397162 cites W2981540341 @default.
- W3202397162 cites W2985406498 @default.
- W3202397162 cites W2995910740 @default.
- W3202397162 cites W3034481430 @default.
- W3202397162 cites W3034562924 @default.
- W3202397162 cites W3035236545 @default.
- W3202397162 cites W3035564946 @default.
- W3202397162 cites W3101956722 @default.
- W3202397162 cites W3106518474 @default.
- W3202397162 cites W3110470025 @default.
- W3202397162 cites W3169545167 @default.
- W3202397162 doi "https://doi.org/10.48550/arxiv.2109.14196" @default.
- W3202397162 hasPublicationYear "2021" @default.
- W3202397162 type Work @default.
- W3202397162 sameAs 3202397162 @default.
- W3202397162 citedByCount "0" @default.
- W3202397162 crossrefType "posted-content" @default.
- W3202397162 hasAuthorship W3202397162A5009948118 @default.
- W3202397162 hasAuthorship W3202397162A5049963367 @default.
- W3202397162 hasAuthorship W3202397162A5051469345 @default.
- W3202397162 hasAuthorship W3202397162A5060343759 @default.
- W3202397162 hasAuthorship W3202397162A5074668738 @default.
- W3202397162 hasBestOaLocation W32023971621 @default.
- W3202397162 hasConcept C115961682 @default.
- W3202397162 hasConcept C119857082 @default.
- W3202397162 hasConcept C134306372 @default.
- W3202397162 hasConcept C154945302 @default.
- W3202397162 hasConcept C165696696 @default.
- W3202397162 hasConcept C177148314 @default.
- W3202397162 hasConcept C33923547 @default.
- W3202397162 hasConcept C36503486 @default.
- W3202397162 hasConcept C38652104 @default.
- W3202397162 hasConcept C41008148 @default.
- W3202397162 hasConcept C89600930 @default.
- W3202397162 hasConceptScore W3202397162C115961682 @default.
- W3202397162 hasConceptScore W3202397162C119857082 @default.
- W3202397162 hasConceptScore W3202397162C134306372 @default.
- W3202397162 hasConceptScore W3202397162C154945302 @default.
- W3202397162 hasConceptScore W3202397162C165696696 @default.
- W3202397162 hasConceptScore W3202397162C177148314 @default.
- W3202397162 hasConceptScore W3202397162C33923547 @default.
- W3202397162 hasConceptScore W3202397162C36503486 @default.
- W3202397162 hasConceptScore W3202397162C38652104 @default.
- W3202397162 hasConceptScore W3202397162C41008148 @default.
- W3202397162 hasConceptScore W3202397162C89600930 @default.
- W3202397162 hasLocation W32023971621 @default.
- W3202397162 hasOpenAccess W3202397162 @default.
- W3202397162 hasPrimaryLocation W32023971621 @default.
- W3202397162 hasRelatedWork W2964604098 @default.
- W3202397162 hasRelatedWork W2987403800 @default.
- W3202397162 hasRelatedWork W2989932438 @default.
- W3202397162 hasRelatedWork W3117155374 @default.
- W3202397162 hasRelatedWork W4224306365 @default.
- W3202397162 hasRelatedWork W4287549950 @default.
- W3202397162 hasRelatedWork W4297619395 @default.
- W3202397162 hasRelatedWork W4300049150 @default.
- W3202397162 hasRelatedWork W4302012741 @default.
- W3202397162 hasRelatedWork W4313477600 @default.