FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3124819632> ?p ?o ?g. }

• W3124819632 abstract "This paper addresses the problem of model compression via knowledge distillation. We advocate for a method that optimizes the output feature of the penultimate layer of the student network and hence is directly related to representation learning. Previous distillation methods which typically impose direct feature matching between the student and the teacher do not take into account the classification problem at hand. On the contrary, our distillation method decouples representation learning and classification and utilizes the teacher's pre-trained classifier to train the student's penultimate layer feature. In particular, for the same input image, we wish the teacher's and student's feature to produce the same output when passed through the teacher's classifier which is achieved with a simple L2 loss. Our method is extremely simple to implement and straightforward to train and is shown to consistently outperform previous state-of-the-art methods over a large set of experimental settings including different (a) network architectures, (b) teacher-student capacities, (c) datasets, and (d) domains." @default.
• W3124819632 created "2021-02-01" @default.
• W3124819632 creator A5024224610 @default.
• W3124819632 creator A5051012442 @default.
• W3124819632 creator A5072822225 @default.
• W3124819632 creator A5081233863 @default.
• W3124819632 date "2021-05-03" @default.
• W3124819632 modified "2023-09-23" @default.
• W3124819632 title "Knowledge distillation via softmax regression representation learning" @default.
• W3124819632 cites W1532325895 @default.
• W3124819632 cites W1821462560 @default.
• W3124819632 cites W2118858186 @default.
• W3124819632 cites W2194775991 @default.
• W3124819632 cites W2233116163 @default.
• W3124819632 cites W2300242332 @default.
• W3124819632 cites W2553303224 @default.
• W3124819632 cites W2561238782 @default.
• W3124819632 cites W2603777577 @default.
• W3124819632 cites W2612445135 @default.
• W3124819632 cites W2731516819 @default.
• W3124819632 cites W2739160792 @default.
• W3124819632 cites W2739879705 @default.
• W3124819632 cites W2796438033 @default.
• W3124819632 cites W2886756692 @default.
• W3124819632 cites W2899771611 @default.
• W3124819632 cites W2951513317 @default.
• W3124819632 cites W2955192706 @default.
• W3124819632 cites W2962930448 @default.
• W3124819632 cites W2963140444 @default.
• W3124819632 cites W2963534679 @default.
• W3124819632 cites W2964111476 @default.
• W3124819632 cites W2964137095 @default.
• W3124819632 cites W2964299589 @default.
• W3124819632 cites W2982157312 @default.
• W3124819632 cites W2982242214 @default.
• W3124819632 cites W2986015886 @default.
• W3124819632 cites W2995197345 @default.
• W3124819632 cites W2995607862 @default.
• W3124819632 cites W3004127093 @default.
• W3124819632 cites W3005680577 @default.
• W3124819632 cites W3021187529 @default.
• W3124819632 cites W3034528892 @default.
• W3124819632 cites W3035163969 @default.
• W3124819632 cites W3035524453 @default.
• W3124819632 cites W3091981646 @default.
• W3124819632 cites W3099057664 @default.
• W3124819632 cites W3108075360 @default.
• W3124819632 cites W3118608800 @default.
• W3124819632 cites W566555209 @default.
• W3124819632 hasPublicationYear "2021" @default.
• W3124819632 type Work @default.
• W3124819632 sameAs 3124819632 @default.
• W3124819632 citedByCount "5" @default.
• W3124819632 countsByYear W31248196322020 @default.
• W3124819632 countsByYear W31248196322021 @default.
• W3124819632 crossrefType "proceedings-article" @default.
• W3124819632 hasAuthorship W3124819632A5024224610 @default.
• W3124819632 hasAuthorship W3124819632A5051012442 @default.
• W3124819632 hasAuthorship W3124819632A5072822225 @default.
• W3124819632 hasAuthorship W3124819632A5081233863 @default.
• W3124819632 hasConcept C111472728 @default.
• W3124819632 hasConcept C119857082 @default.
• W3124819632 hasConcept C138885662 @default.
• W3124819632 hasConcept C153180895 @default.
• W3124819632 hasConcept C154945302 @default.
• W3124819632 hasConcept C17744445 @default.
• W3124819632 hasConcept C178790620 @default.
• W3124819632 hasConcept C185592680 @default.
• W3124819632 hasConcept C188441871 @default.
• W3124819632 hasConcept C199539241 @default.
• W3124819632 hasConcept C204030448 @default.
• W3124819632 hasConcept C2776359362 @default.
• W3124819632 hasConcept C2776401178 @default.
• W3124819632 hasConcept C2780586882 @default.
• W3124819632 hasConcept C41008148 @default.
• W3124819632 hasConcept C41895202 @default.
• W3124819632 hasConcept C50644808 @default.
• W3124819632 hasConcept C52622490 @default.
• W3124819632 hasConcept C59404180 @default.
• W3124819632 hasConcept C94625758 @default.
• W3124819632 hasConcept C95623464 @default.
• W3124819632 hasConceptScore W3124819632C111472728 @default.
• W3124819632 hasConceptScore W3124819632C119857082 @default.
• W3124819632 hasConceptScore W3124819632C138885662 @default.
• W3124819632 hasConceptScore W3124819632C153180895 @default.
• W3124819632 hasConceptScore W3124819632C154945302 @default.
• W3124819632 hasConceptScore W3124819632C17744445 @default.
• W3124819632 hasConceptScore W3124819632C178790620 @default.
• W3124819632 hasConceptScore W3124819632C185592680 @default.
• W3124819632 hasConceptScore W3124819632C188441871 @default.
• W3124819632 hasConceptScore W3124819632C199539241 @default.
• W3124819632 hasConceptScore W3124819632C204030448 @default.
• W3124819632 hasConceptScore W3124819632C2776359362 @default.
• W3124819632 hasConceptScore W3124819632C2776401178 @default.
• W3124819632 hasConceptScore W3124819632C2780586882 @default.
• W3124819632 hasConceptScore W3124819632C41008148 @default.
• W3124819632 hasConceptScore W3124819632C41895202 @default.
• W3124819632 hasConceptScore W3124819632C50644808 @default.
• W3124819632 hasConceptScore W3124819632C52622490 @default.
• W3124819632 hasConceptScore W3124819632C59404180 @default.

• next