FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3105676370> ?p ?o ?g. }

• W3105676370 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.
• W3105676370 created "2020-11-23" @default.
• W3105676370 creator A5030456600 @default.
• W3105676370 creator A5054356243 @default.
• W3105676370 creator A5060490339 @default.
• W3105676370 date "2020-11-09" @default.
• W3105676370 modified "2023-09-27" @default.
• W3105676370 title "Improving Classifier Confidence using Lossy Label-Invariant Transformations" @default.
• W3105676370 cites W1618905105 @default.
• W3105676370 cites W1757796397 @default.
• W3105676370 cites W2012942264 @default.
• W3105676370 cites W2073241381 @default.
• W3105676370 cites W2108598243 @default.
• W3105676370 cites W2112796928 @default.
• W3105676370 cites W2143612262 @default.
• W3105676370 cites W2194775991 @default.
• W3105676370 cites W2254249950 @default.
• W3105676370 cites W2331143823 @default.
• W3105676370 cites W2401231614 @default.
• W3105676370 cites W2549139847 @default.
• W3105676370 cites W2626967530 @default.
• W3105676370 cites W2766447205 @default.
• W3105676370 cites W2768927604 @default.
• W3105676370 cites W2774412855 @default.
• W3105676370 cites W2804697534 @default.
• W3105676370 cites W2807776404 @default.
• W3105676370 cites W2901954625 @default.
• W3105676370 cites W2921861056 @default.
• W3105676370 cites W2945017166 @default.
• W3105676370 cites W2950300355 @default.
• W3105676370 cites W2963036823 @default.
• W3105676370 cites W2963238274 @default.
• W3105676370 cites W2963446712 @default.
• W3105676370 cites W2964059111 @default.
• W3105676370 cites W2970121940 @default.
• W3105676370 cites W2970206392 @default.
• W3105676370 cites W2971118045 @default.
• W3105676370 cites W2982509564 @default.
• W3105676370 cites W2995851819 @default.
• W3105676370 cites W3008912312 @default.
• W3105676370 cites W3011253634 @default.
• W3105676370 cites W3034905393 @default.
• W3105676370 cites W3036157921 @default.
• W3105676370 cites W3037275730 @default.
• W3105676370 cites W3038321616 @default.
• W3105676370 cites W3118608800 @default.
• W3105676370 hasPublicationYear "2020" @default.
• W3105676370 type Work @default.
• W3105676370 sameAs 3105676370 @default.
• W3105676370 citedByCount "1" @default.
• W3105676370 countsByYear W31056763702021 @default.
• W3105676370 crossrefType "posted-content" @default.
• W3105676370 hasAuthorship W3105676370A5030456600 @default.
• W3105676370 hasAuthorship W3105676370A5054356243 @default.
• W3105676370 hasAuthorship W3105676370A5060490339 @default.
• W3105676370 hasConcept C105795698 @default.
• W3105676370 hasConcept C119857082 @default.
• W3105676370 hasConcept C153083717 @default.
• W3105676370 hasConcept C153180895 @default.
• W3105676370 hasConcept C154945302 @default.
• W3105676370 hasConcept C165021410 @default.
• W3105676370 hasConcept C165838908 @default.
• W3105676370 hasConcept C190470478 @default.
• W3105676370 hasConcept C33923547 @default.
• W3105676370 hasConcept C37914503 @default.
• W3105676370 hasConcept C41008148 @default.
• W3105676370 hasConcept C95623464 @default.
• W3105676370 hasConceptScore W3105676370C105795698 @default.
• W3105676370 hasConceptScore W3105676370C119857082 @default.
• W3105676370 hasConceptScore W3105676370C153083717 @default.
• W3105676370 hasConceptScore W3105676370C153180895 @default.
• W3105676370 hasConceptScore W3105676370C154945302 @default.
• W3105676370 hasConceptScore W3105676370C165021410 @default.
• W3105676370 hasConceptScore W3105676370C165838908 @default.
• W3105676370 hasConceptScore W3105676370C190470478 @default.
• W3105676370 hasConceptScore W3105676370C33923547 @default.
• W3105676370 hasConceptScore W3105676370C37914503 @default.
• W3105676370 hasConceptScore W3105676370C41008148 @default.
• W3105676370 hasConceptScore W3105676370C95623464 @default.
• W3105676370 hasLocation W31056763701 @default.
• W3105676370 hasOpenAccess W3105676370 @default.
• W3105676370 hasPrimaryLocation W31056763701 @default.
• W3105676370 hasRelatedWork W2804869193 @default.
• W3105676370 hasRelatedWork W2810438149 @default.
• W3105676370 hasRelatedWork W2899920074 @default.
• W3105676370 hasRelatedWork W2914277910 @default.
• W3105676370 hasRelatedWork W2918288062 @default.
• W3105676370 hasRelatedWork W2920830204 @default.
• W3105676370 hasRelatedWork W2970388146 @default.
• W3105676370 hasRelatedWork W3004854517 @default.
• W3105676370 hasRelatedWork W3015210400 @default.
• W3105676370 hasRelatedWork W3021147025 @default.
• W3105676370 hasRelatedWork W3042865207 @default.
• W3105676370 hasRelatedWork W3080854194 @default.
• W3105676370 hasRelatedWork W3082390656 @default.
• W3105676370 hasRelatedWork W3083184539 @default.
• W3105676370 hasRelatedWork W3090855199 @default.
• W3105676370 hasRelatedWork W3101812930 @default.
• W3105676370 hasRelatedWork W3116425452 @default.
• W3105676370 hasRelatedWork W3120549604 @default.

• next