Matches in SemOpenAlex for { <https://semopenalex.org/work/W3119583506> ?p ?o ?g. }
- W3119583506 abstract "Learning with labels noise has gained significant traction recently due to the sensitivity of deep neural networks under label noise under common loss functions. Losses that are theoretically robust to label noise, however, often makes training difficult. Consequently, several recently proposed methods, such as Meta-Weight-Net (MW-Net), use a small number of unbiased, clean samples to learn a weighting function that downweights samples that are likely to have corrupted labels under the meta-learning framework. However, obtaining such a set of clean samples is not always feasible in practice. In this paper, we analytically show that one can easily train MW-Net without access to clean samples simply by using a loss function that is robust to label noise, such as mean absolute error, as the meta objective to train the weighting network. We experimentally show that our method beats all existing methods that do not use clean samples and performs on-par with methods that use gold samples on benchmark datasets across various noise types and noise rates." @default.
- W3119583506 created "2021-01-18" @default.
- W3119583506 creator A5003881665 @default.
- W3119583506 creator A5063813962 @default.
- W3119583506 date "2021-01-01" @default.
- W3119583506 modified "2023-10-13" @default.
- W3119583506 title "Do We Really Need Gold Samples for Sample Weighting under Label Noise?" @default.
- W3119583506 cites W1975128126 @default.
- W3119583506 cites W1982032418 @default.
- W3119583506 cites W1994550352 @default.
- W3119583506 cites W1994616650 @default.
- W3119583506 cites W2124219775 @default.
- W3119583506 cites W2167460663 @default.
- W3119583506 cites W2194775991 @default.
- W3119583506 cites W2296073425 @default.
- W3119583506 cites W2577784528 @default.
- W3119583506 cites W2592335154 @default.
- W3119583506 cites W2801602507 @default.
- W3119583506 cites W2905290040 @default.
- W3119583506 cites W2948606739 @default.
- W3119583506 cites W2962762068 @default.
- W3119583506 cites W2963351448 @default.
- W3119583506 cites W2963859210 @default.
- W3119583506 cites W2964155802 @default.
- W3119583506 cites W2964274690 @default.
- W3119583506 cites W2964292098 @default.
- W3119583506 cites W2967052791 @default.
- W3119583506 cites W2967637014 @default.
- W3119583506 cites W2978625989 @default.
- W3119583506 cites W2981873476 @default.
- W3119583506 cites W2981952612 @default.
- W3119583506 cites W2988966271 @default.
- W3119583506 cites W3009953871 @default.
- W3119583506 cites W3035716296 @default.
- W3119583506 cites W3100570787 @default.
- W3119583506 cites W3102825217 @default.
- W3119583506 doi "https://doi.org/10.1109/wacv48630.2021.00397" @default.
- W3119583506 hasPublicationYear "2021" @default.
- W3119583506 type Work @default.
- W3119583506 sameAs 3119583506 @default.
- W3119583506 citedByCount "7" @default.
- W3119583506 countsByYear W31195835062021 @default.
- W3119583506 countsByYear W31195835062022 @default.
- W3119583506 countsByYear W31195835062023 @default.
- W3119583506 crossrefType "proceedings-article" @default.
- W3119583506 hasAuthorship W3119583506A5003881665 @default.
- W3119583506 hasAuthorship W3119583506A5063813962 @default.
- W3119583506 hasBestOaLocation W31195835062 @default.
- W3119583506 hasConcept C115961682 @default.
- W3119583506 hasConcept C119857082 @default.
- W3119583506 hasConcept C121332964 @default.
- W3119583506 hasConcept C124101348 @default.
- W3119583506 hasConcept C13280743 @default.
- W3119583506 hasConcept C14036430 @default.
- W3119583506 hasConcept C153180895 @default.
- W3119583506 hasConcept C154945302 @default.
- W3119583506 hasConcept C163294075 @default.
- W3119583506 hasConcept C177264268 @default.
- W3119583506 hasConcept C183115368 @default.
- W3119583506 hasConcept C185592680 @default.
- W3119583506 hasConcept C185798385 @default.
- W3119583506 hasConcept C198531522 @default.
- W3119583506 hasConcept C199360897 @default.
- W3119583506 hasConcept C205649164 @default.
- W3119583506 hasConcept C24890656 @default.
- W3119583506 hasConcept C29265498 @default.
- W3119583506 hasConcept C41008148 @default.
- W3119583506 hasConcept C43617362 @default.
- W3119583506 hasConcept C50644808 @default.
- W3119583506 hasConcept C70136482 @default.
- W3119583506 hasConcept C78458016 @default.
- W3119583506 hasConcept C86803240 @default.
- W3119583506 hasConcept C99498987 @default.
- W3119583506 hasConceptScore W3119583506C115961682 @default.
- W3119583506 hasConceptScore W3119583506C119857082 @default.
- W3119583506 hasConceptScore W3119583506C121332964 @default.
- W3119583506 hasConceptScore W3119583506C124101348 @default.
- W3119583506 hasConceptScore W3119583506C13280743 @default.
- W3119583506 hasConceptScore W3119583506C14036430 @default.
- W3119583506 hasConceptScore W3119583506C153180895 @default.
- W3119583506 hasConceptScore W3119583506C154945302 @default.
- W3119583506 hasConceptScore W3119583506C163294075 @default.
- W3119583506 hasConceptScore W3119583506C177264268 @default.
- W3119583506 hasConceptScore W3119583506C183115368 @default.
- W3119583506 hasConceptScore W3119583506C185592680 @default.
- W3119583506 hasConceptScore W3119583506C185798385 @default.
- W3119583506 hasConceptScore W3119583506C198531522 @default.
- W3119583506 hasConceptScore W3119583506C199360897 @default.
- W3119583506 hasConceptScore W3119583506C205649164 @default.
- W3119583506 hasConceptScore W3119583506C24890656 @default.
- W3119583506 hasConceptScore W3119583506C29265498 @default.
- W3119583506 hasConceptScore W3119583506C41008148 @default.
- W3119583506 hasConceptScore W3119583506C43617362 @default.
- W3119583506 hasConceptScore W3119583506C50644808 @default.
- W3119583506 hasConceptScore W3119583506C70136482 @default.
- W3119583506 hasConceptScore W3119583506C78458016 @default.
- W3119583506 hasConceptScore W3119583506C86803240 @default.
- W3119583506 hasConceptScore W3119583506C99498987 @default.
- W3119583506 hasLocation W31195835061 @default.
- W3119583506 hasLocation W31195835062 @default.