FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4319341372> ?p ?o ?g. }

• W4319341372 endingPage "16" @default.
• W4319341372 startingPage "1" @default.
• W4319341372 abstract "Supervised deep learning has achieved tremendous success in many computer vision tasks, which however is prone to overfit noisy labels. To mitigate the undesirable influence of noisy labels, robust loss functions offer a feasible approach to achieve noise-tolerant learning. In this work, we systematically study the problem of noise-tolerant learning with respect to both classification and regression. Specifically, we propose a new class of loss function, namely asymmetric loss functions (ALFs), which are tailored to satisfy the Bayes-optimal condition and thus are robust to noisy labels. For classification, we investigate general theoretical properties of ALFs on categorical noisy labels, and introduce the asymmetry ratio to measure the asymmetry of a loss function. We extend several commonly-used loss functions, and establish the necessary and sufficient conditions to make them asymmetric and thus noise-tolerant. For regression, we extend the concept of noise-tolerant learning for image restoration with continuous noisy labels. We theoretically prove that lp loss ( ) is noise-tolerant for targets with the additive white Gaussian noise. For targets with general noise, we introduce two losses as surrogates of l0 loss that seeks the mode when clean pixels keep dominant. Experimental results demonstrate that ALFs can achieve better or comparative performance compared with the state-of-the-arts. The source code of our method is available at: https://github.com/hitcszx/ALFs." @default.
• W4319341372 created "2023-02-08" @default.
• W4319341372 creator A5024401174 @default.
• W4319341372 creator A5025244867 @default.
• W4319341372 creator A5029696721 @default.
• W4319341372 creator A5086046733 @default.
• W4319341372 creator A5087165831 @default.
• W4319341372 date "2023-01-01" @default.
• W4319341372 modified "2023-10-18" @default.
• W4319341372 title "Asymmetric Loss Functions for Noise-Tolerant Learning: Theory and Applications" @default.
• W4319341372 cites W1514928307 @default.
• W4319341372 cites W1906770428 @default.
• W4319341372 cites W1975128126 @default.
• W4319341372 cites W1982032418 @default.
• W4319341372 cites W2048695508 @default.
• W4319341372 cites W2056370875 @default.
• W4319341372 cites W2097073572 @default.
• W4319341372 cites W2102348129 @default.
• W4319341372 cites W2103496339 @default.
• W4319341372 cites W2108598243 @default.
• W4319341372 cites W2112796928 @default.
• W4319341372 cites W2130184048 @default.
• W4319341372 cites W2161149720 @default.
• W4319341372 cites W2194775991 @default.
• W4319341372 cites W2508457857 @default.
• W4319341372 cites W2610690067 @default.
• W4319341372 cites W2902857081 @default.
• W4319341372 cites W2962762068 @default.
• W4319341372 cites W2963351448 @default.
• W4319341372 cites W2963697299 @default.
• W4319341372 cites W2981873476 @default.
• W4319341372 cites W2981952612 @default.
• W4319341372 cites W3034902810 @default.
• W4319341372 cites W3035542568 @default.
• W4319341372 cites W3047103180 @default.
• W4319341372 cites W3065974826 @default.
• W4319341372 cites W3097262261 @default.
• W4319341372 cites W3178192988 @default.
• W4319341372 cites W4214585162 @default.
• W4319341372 cites W4221161368 @default.
• W4319341372 doi "https://doi.org/10.1109/tpami.2023.3236459" @default.
• W4319341372 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37022833" @default.
• W4319341372 hasPublicationYear "2023" @default.
• W4319341372 type Work @default.
• W4319341372 citedByCount "2" @default.
• W4319341372 countsByYear W43193413722023 @default.
• W4319341372 crossrefType "journal-article" @default.
• W4319341372 hasAuthorship W4319341372A5024401174 @default.
• W4319341372 hasAuthorship W4319341372A5025244867 @default.
• W4319341372 hasAuthorship W4319341372A5029696721 @default.
• W4319341372 hasAuthorship W4319341372A5086046733 @default.
• W4319341372 hasAuthorship W4319341372A5087165831 @default.
• W4319341372 hasConcept C111919701 @default.
• W4319341372 hasConcept C115961682 @default.
• W4319341372 hasConcept C119857082 @default.
• W4319341372 hasConcept C14036430 @default.
• W4319341372 hasConcept C153180895 @default.
• W4319341372 hasConcept C154945302 @default.
• W4319341372 hasConcept C163294075 @default.
• W4319341372 hasConcept C22019652 @default.
• W4319341372 hasConcept C29265498 @default.
• W4319341372 hasConcept C41008148 @default.
• W4319341372 hasConcept C4199805 @default.
• W4319341372 hasConcept C43126263 @default.
• W4319341372 hasConcept C50644808 @default.
• W4319341372 hasConcept C78458016 @default.
• W4319341372 hasConcept C86803240 @default.
• W4319341372 hasConcept C99498987 @default.
• W4319341372 hasConceptScore W4319341372C111919701 @default.
• W4319341372 hasConceptScore W4319341372C115961682 @default.
• W4319341372 hasConceptScore W4319341372C119857082 @default.
• W4319341372 hasConceptScore W4319341372C14036430 @default.
• W4319341372 hasConceptScore W4319341372C153180895 @default.
• W4319341372 hasConceptScore W4319341372C154945302 @default.
• W4319341372 hasConceptScore W4319341372C163294075 @default.
• W4319341372 hasConceptScore W4319341372C22019652 @default.
• W4319341372 hasConceptScore W4319341372C29265498 @default.
• W4319341372 hasConceptScore W4319341372C41008148 @default.
• W4319341372 hasConceptScore W4319341372C4199805 @default.
• W4319341372 hasConceptScore W4319341372C43126263 @default.
• W4319341372 hasConceptScore W4319341372C50644808 @default.
• W4319341372 hasConceptScore W4319341372C78458016 @default.
• W4319341372 hasConceptScore W4319341372C86803240 @default.
• W4319341372 hasConceptScore W4319341372C99498987 @default.
• W4319341372 hasFunder F4320321001 @default.
• W4319341372 hasLocation W43193413721 @default.
• W4319341372 hasLocation W43193413722 @default.
• W4319341372 hasOpenAccess W4319341372 @default.
• W4319341372 hasPrimaryLocation W43193413721 @default.
• W4319341372 hasRelatedWork W1996541855 @default.
• W4319341372 hasRelatedWork W2115137940 @default.
• W4319341372 hasRelatedWork W2767651786 @default.
• W4319341372 hasRelatedWork W2989932438 @default.
• W4319341372 hasRelatedWork W3011996705 @default.
• W4319341372 hasRelatedWork W3099765033 @default.
• W4319341372 hasRelatedWork W3175189414 @default.
• W4319341372 hasRelatedWork W3206592002 @default.
• W4319341372 hasRelatedWork W4210794429 @default.

• next