Matches in SemOpenAlex for { <https://semopenalex.org/work/W3159142482> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W3159142482 endingPage "4059" @default.
- W3159142482 startingPage "4051" @default.
- W3159142482 abstract "Providing reliable model uncertainty estimates is imperative to enabling robust decision making by autonomous agents and humans alike. While recently there have been significant advances in confidence calibration for trained models, examples with poor calibration persist in most calibrated models. Consequently, multiple techniques have been proposed that leverage label-invariant transformations of the input (i.e., an input manifold) to improve worst-case confidence calibration. However, manifold-based confidence calibration techniques generally do not scale and/or require expensive retraining when applied to models with large input spaces (e.g., ImageNet). In this paper, we present the recursive lossy label-invariant calibration (ReCal) technique that leverages label-invariant transformations of the input that induce a loss of discriminatory information to recursively group (and calibrate) inputs - without requiring model retraining. We show that ReCal outperforms other calibration methods on multiple datasets, especially, on large-scale datasets such as ImageNet." @default.
- W3159142482 created "2021-05-10" @default.
- W3159142482 creator A5030456600 @default.
- W3159142482 creator A5054356243 @default.
- W3159142482 creator A5060490339 @default.
- W3159142482 date "2021-03-18" @default.
- W3159142482 modified "2023-09-23" @default.
- W3159142482 title "Improving Classifier Confidence using Lossy Label-Invariant Transformations" @default.
- W3159142482 hasPublicationYear "2021" @default.
- W3159142482 type Work @default.
- W3159142482 sameAs 3159142482 @default.
- W3159142482 citedByCount "0" @default.
- W3159142482 crossrefType "proceedings-article" @default.
- W3159142482 hasAuthorship W3159142482A5030456600 @default.
- W3159142482 hasAuthorship W3159142482A5054356243 @default.
- W3159142482 hasAuthorship W3159142482A5060490339 @default.
- W3159142482 hasConcept C105795698 @default.
- W3159142482 hasConcept C119857082 @default.
- W3159142482 hasConcept C153083717 @default.
- W3159142482 hasConcept C153180895 @default.
- W3159142482 hasConcept C154945302 @default.
- W3159142482 hasConcept C165021410 @default.
- W3159142482 hasConcept C165838908 @default.
- W3159142482 hasConcept C190470478 @default.
- W3159142482 hasConcept C33923547 @default.
- W3159142482 hasConcept C37914503 @default.
- W3159142482 hasConcept C41008148 @default.
- W3159142482 hasConcept C95623464 @default.
- W3159142482 hasConceptScore W3159142482C105795698 @default.
- W3159142482 hasConceptScore W3159142482C119857082 @default.
- W3159142482 hasConceptScore W3159142482C153083717 @default.
- W3159142482 hasConceptScore W3159142482C153180895 @default.
- W3159142482 hasConceptScore W3159142482C154945302 @default.
- W3159142482 hasConceptScore W3159142482C165021410 @default.
- W3159142482 hasConceptScore W3159142482C165838908 @default.
- W3159142482 hasConceptScore W3159142482C190470478 @default.
- W3159142482 hasConceptScore W3159142482C33923547 @default.
- W3159142482 hasConceptScore W3159142482C37914503 @default.
- W3159142482 hasConceptScore W3159142482C41008148 @default.
- W3159142482 hasConceptScore W3159142482C95623464 @default.
- W3159142482 hasLocation W31591424821 @default.
- W3159142482 hasOpenAccess W3159142482 @default.
- W3159142482 hasPrimaryLocation W31591424821 @default.
- W3159142482 hasRelatedWork W2168920183 @default.
- W3159142482 hasRelatedWork W2431080869 @default.
- W3159142482 hasRelatedWork W2606711537 @default.
- W3159142482 hasRelatedWork W2618238855 @default.
- W3159142482 hasRelatedWork W2763270288 @default.
- W3159142482 hasRelatedWork W2795282075 @default.
- W3159142482 hasRelatedWork W2905290040 @default.
- W3159142482 hasRelatedWork W2951236680 @default.
- W3159142482 hasRelatedWork W2963134708 @default.
- W3159142482 hasRelatedWork W2969594481 @default.
- W3159142482 hasRelatedWork W2976181893 @default.
- W3159142482 hasRelatedWork W2996463194 @default.
- W3159142482 hasRelatedWork W3010165109 @default.
- W3159142482 hasRelatedWork W3092256610 @default.
- W3159142482 hasRelatedWork W3098898971 @default.
- W3159142482 hasRelatedWork W3112282987 @default.
- W3159142482 hasRelatedWork W3120587221 @default.
- W3159142482 hasRelatedWork W3121713357 @default.
- W3159142482 hasRelatedWork W3168719605 @default.
- W3159142482 hasRelatedWork W3193615289 @default.
- W3159142482 hasVolume "130" @default.
- W3159142482 isParatext "false" @default.
- W3159142482 isRetracted "false" @default.
- W3159142482 magId "3159142482" @default.
- W3159142482 workType "article" @default.