FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385151915> ?p ?o ?g. }

• W4385151915 endingPage "2987" @default.
• W4385151915 startingPage "2974" @default.
• W4385151915 abstract "Knowledge distillation (KD), best known as an effective method for model compression, aims at transferring the knowledge of a bigger network (teacher) to a much smaller network (student). Conventional KD methods usually employ the teacher model trained in a supervised manner, where output labels are treated only as targets. Extending this supervised scheme further, we introduce a new type of teacher model for connectionist temporal classification (CTC)-based sequence models, namely Oracle Teacher, that leverages both the source inputs and the output labels as the teacher model's input. Since the Oracle Teacher learns a more accurate CTC alignment by referring to the target information, it can provide the student with more optimal guidance. One potential risk for the proposed approach is a trivial solution that the model's output directly copies the target input. Based on a many-to-one mapping property of the CTC algorithm, we present a training strategy that can effectively prevent the trivial solution and thus enables utilizing both source and target inputs for model training. Extensive experiments are conducted on two sequence learning tasks: speech recognition and scene text recognition. From the experimental results, we empirically show that the proposed model improves the students across these tasks while achieving a considerable speed-up in the teacher model's training time." @default.
• W4385151915 created "2023-07-23" @default.
• W4385151915 creator A5016783920 @default.
• W4385151915 creator A5051356511 @default.
• W4385151915 creator A5068972361 @default.
• W4385151915 creator A5073477530 @default.
• W4385151915 creator A5074660905 @default.
• W4385151915 date "2023-01-01" @default.
• W4385151915 modified "2023-09-25" @default.
• W4385151915 title "Oracle Teacher: Leveraging Target Information for Better Knowledge Distillation of CTC Models" @default.
• W4385151915 cites W1494198834 @default.
• W4385151915 cites W1533416326 @default.
• W4385151915 cites W1971822075 @default.
• W4385151915 cites W1981283549 @default.
• W4385151915 cites W2008806374 @default.
• W4385151915 cites W2127141656 @default.
• W4385151915 cites W2140132917 @default.
• W4385151915 cites W2144554289 @default.
• W4385151915 cites W2146835493 @default.
• W4385151915 cites W2194187530 @default.
• W4385151915 cites W2233116163 @default.
• W4385151915 cites W2294370754 @default.
• W4385151915 cites W2343052201 @default.
• W4385151915 cites W2402040300 @default.
• W4385151915 cites W2507699225 @default.
• W4385151915 cites W2508418541 @default.
• W4385151915 cites W2711861986 @default.
• W4385151915 cites W2739879705 @default.
• W4385151915 cites W2747909401 @default.
• W4385151915 cites W2809273748 @default.
• W4385151915 cites W2889129739 @default.
• W4385151915 cites W2892008152 @default.
• W4385151915 cites W2913178639 @default.
• W4385151915 cites W2936993002 @default.
• W4385151915 cites W2938974662 @default.
• W4385151915 cites W2962784628 @default.
• W4385151915 cites W2963140444 @default.
• W4385151915 cites W2963864497 @default.
• W4385151915 cites W2963920996 @default.
• W4385151915 cites W2964110616 @default.
• W4385151915 cites W2964111476 @default.
• W4385151915 cites W2973040747 @default.
• W4385151915 cites W2973215447 @default.
• W4385151915 cites W2997006708 @default.
• W4385151915 cites W3004127093 @default.
• W4385151915 cites W3015537910 @default.
• W4385151915 cites W3015565442 @default.
• W4385151915 cites W3015752032 @default.
• W4385151915 cites W3035402938 @default.
• W4385151915 cites W3097777922 @default.
• W4385151915 cites W3127215335 @default.
• W4385151915 cites W3155427814 @default.
• W4385151915 cites W3156828761 @default.
• W4385151915 cites W3162833755 @default.
• W4385151915 cites W3163044982 @default.
• W4385151915 cites W3177196641 @default.
• W4385151915 cites W3179897446 @default.
• W4385151915 cites W3200601846 @default.
• W4385151915 cites W3200868928 @default.
• W4385151915 cites W3203140070 @default.
• W4385151915 cites W4224821750 @default.
• W4385151915 cites W4224918069 @default.
• W4385151915 cites W4225308107 @default.
• W4385151915 cites W4226120743 @default.
• W4385151915 cites W4226521565 @default.
• W4385151915 cites W4297841557 @default.
• W4385151915 cites W4319862663 @default.
• W4385151915 doi "https://doi.org/10.1109/taslp.2023.3297955" @default.
• W4385151915 hasPublicationYear "2023" @default.
• W4385151915 type Work @default.
• W4385151915 citedByCount "0" @default.
• W4385151915 crossrefType "journal-article" @default.
• W4385151915 hasAuthorship W4385151915A5016783920 @default.
• W4385151915 hasAuthorship W4385151915A5051356511 @default.
• W4385151915 hasAuthorship W4385151915A5068972361 @default.
• W4385151915 hasAuthorship W4385151915A5073477530 @default.
• W4385151915 hasAuthorship W4385151915A5074660905 @default.
• W4385151915 hasBestOaLocation W43851519152 @default.
• W4385151915 hasConcept C111472728 @default.
• W4385151915 hasConcept C115903868 @default.
• W4385151915 hasConcept C119857082 @default.
• W4385151915 hasConcept C134306372 @default.
• W4385151915 hasConcept C138885662 @default.
• W4385151915 hasConcept C154945302 @default.
• W4385151915 hasConcept C178790620 @default.
• W4385151915 hasConcept C185592680 @default.
• W4385151915 hasConcept C189950617 @default.
• W4385151915 hasConcept C204030448 @default.
• W4385151915 hasConcept C2778112365 @default.
• W4385151915 hasConcept C33923547 @default.
• W4385151915 hasConcept C41008148 @default.
• W4385151915 hasConcept C50644808 @default.
• W4385151915 hasConcept C54355233 @default.
• W4385151915 hasConcept C55166926 @default.
• W4385151915 hasConcept C77618280 @default.
• W4385151915 hasConcept C8521452 @default.
• W4385151915 hasConcept C86803240 @default.
• W4385151915 hasConceptScore W4385151915C111472728 @default.

• next