FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3112883392> ?p ?o ?g. }

• W3112883392 abstract "Reliably quantifying the confidence of deep neural classifiers is a challenging yet fundamental requirement for deploying such models in safety-critical applications. In this paper, we introduce a novel target criterion for model confidence, namely the true class probability (TCP). We show that TCP offers better properties for confidence estimation than standard maximum class probability (MCP). Since the true class is by essence unknown at test time, we propose to learn TCP criterion from data with an auxiliary model, introducing a specific learning scheme adapted to this context. We evaluate our approach on the task of failure prediction and of self-training with pseudo-labels for domain adaptation, which both necessitate effective confidence estimates. Extensive experiments are conducted for validating the relevance of the proposed approach in each task. We study various network architectures and experiment with small and large datasets for image classification and semantic segmentation. In every tested benchmark, our approach outperforms strong baselines." @default.
• W3112883392 created "2020-12-21" @default.
• W3112883392 creator A5016868273 @default.
• W3112883392 creator A5017490804 @default.
• W3112883392 creator A5019603184 @default.
• W3112883392 creator A5022871131 @default.
• W3112883392 creator A5063633079 @default.
• W3112883392 creator A5076170578 @default.
• W3112883392 date "2020-12-11" @default.
• W3112883392 modified "2023-09-23" @default.
• W3112883392 title "Confidence Estimation via Auxiliary Models" @default.
• W3112883392 cites W132090532 @default.
• W3112883392 cites W1567512734 @default.
• W3112883392 cites W1614298861 @default.
• W3112883392 cites W1672197616 @default.
• W3112883392 cites W1719489212 @default.
• W3112883392 cites W179875071 @default.
• W3112883392 cites W1821462560 @default.
• W3112883392 cites W1922655562 @default.
• W3112883392 cites W1932198206 @default.
• W3112883392 cites W2101946573 @default.
• W3112883392 cites W2103328396 @default.
• W3112883392 cites W2112796928 @default.
• W3112883392 cites W2122217421 @default.
• W3112883392 cites W2144746247 @default.
• W3112883392 cites W2145494108 @default.
• W3112883392 cites W2147947791 @default.
• W3112883392 cites W2159391028 @default.
• W3112883392 cites W2160815625 @default.
• W3112883392 cites W2161813894 @default.
• W3112883392 cites W2163605009 @default.
• W3112883392 cites W2164411961 @default.
• W3112883392 cites W2169413195 @default.
• W3112883392 cites W2335728318 @default.
• W3112883392 cites W2340897893 @default.
• W3112883392 cites W2412782625 @default.
• W3112883392 cites W2431874326 @default.
• W3112883392 cites W2462906003 @default.
• W3112883392 cites W2487365028 @default.
• W3112883392 cites W2522957395 @default.
• W3112883392 cites W2531327146 @default.
• W3112883392 cites W2555956183 @default.
• W3112883392 cites W2597787948 @default.
• W3112883392 cites W2613718673 @default.
• W3112883392 cites W2781228439 @default.
• W3112883392 cites W2893463703 @default.
• W3112883392 cites W2894319790 @default.
• W3112883392 cites W2895281799 @default.
• W3112883392 cites W2898630520 @default.
• W3112883392 cites W2899771611 @default.
• W3112883392 cites W2908048420 @default.
• W3112883392 cites W2914735679 @default.
• W3112883392 cites W2932414082 @default.
• W3112883392 cites W2951266961 @default.
• W3112883392 cites W2962793481 @default.
• W3112883392 cites W2962808524 @default.
• W3112883392 cites W2963037989 @default.
• W3112883392 cites W2963073217 @default.
• W3112883392 cites W2963107255 @default.
• W3112883392 cites W2963207607 @default.
• W3112883392 cites W2963238274 @default.
• W3112883392 cites W2963351448 @default.
• W3112883392 cites W2963441669 @default.
• W3112883392 cites W2963505832 @default.
• W3112883392 cites W2963613748 @default.
• W3112883392 cites W2963693742 @default.
• W3112883392 cites W2964003311 @default.
• W3112883392 cites W2964049460 @default.
• W3112883392 cites W2964059111 @default.
• W3112883392 cites W2964121744 @default.
• W3112883392 cites W2964153729 @default.
• W3112883392 cites W2964212410 @default.
• W3112883392 cites W2970565181 @default.
• W3112883392 cites W2970971581 @default.
• W3112883392 cites W2971130081 @default.
• W3112883392 cites W2972285644 @default.
• W3112883392 cites W2985406498 @default.
• W3112883392 cites W3034679848 @default.
• W3112883392 cites W3040318838 @default.
• W3112883392 cites W3106250896 @default.
• W3112883392 cites W3118608800 @default.
• W3112883392 hasPublicationYear "2020" @default.
• W3112883392 type Work @default.
• W3112883392 sameAs 3112883392 @default.
• W3112883392 citedByCount "1" @default.
• W3112883392 countsByYear W31128833922021 @default.
• W3112883392 crossrefType "posted-content" @default.
• W3112883392 hasAuthorship W3112883392A5016868273 @default.
• W3112883392 hasAuthorship W3112883392A5017490804 @default.
• W3112883392 hasAuthorship W3112883392A5019603184 @default.
• W3112883392 hasAuthorship W3112883392A5022871131 @default.
• W3112883392 hasAuthorship W3112883392A5063633079 @default.
• W3112883392 hasAuthorship W3112883392A5076170578 @default.
• W3112883392 hasConcept C105795698 @default.
• W3112883392 hasConcept C119857082 @default.
• W3112883392 hasConcept C124101348 @default.
• W3112883392 hasConcept C127413603 @default.
• W3112883392 hasConcept C13280743 @default.
• W3112883392 hasConcept C151730666 @default.
• W3112883392 hasConcept C154945302 @default.

• next