Matches in SemOpenAlex for { <https://semopenalex.org/work/W3200103931> ?p ?o ?g. }
- W3200103931 abstract "Deep learning models are proven to deliver satisfactory results on training a complex non-linear relationship between the set of input features and different task outputs. However, they are memory intensive and require good computational power for both training as well as inferencing. In literature one can find different model compression techniques which enables easy deployment on edge devices. Knowledge distillation is one such approach where the knowledge of complex teacher model is transferred to a lower parameter student model. However, the limitation is that the architecture of the student model should be comparable to the complex teacher model for better knowledge transfer. Due to this limitation, we cannot deploy this student model that learns from a complex and huge teacher on edge devices. In this work, we propose to use a combined student approach wherein different student models learn from a common teacher model. Further, we propose a unique loss function which will train multiple student models simultaneously. An advantage of this approach is that these student models can be as simple as possible when compared with traditional single student model and also the complex teacher model. Finally, we provide an extensive evaluation to prove that our approach can improve the overall accuracy significantly and allow a further compression by 10% when compared with generic model." @default.
- W3200103931 created "2021-09-27" @default.
- W3200103931 creator A5019643071 @default.
- W3200103931 creator A5023878418 @default.
- W3200103931 creator A5041130310 @default.
- W3200103931 creator A5060733879 @default.
- W3200103931 date "2021-07-27" @default.
- W3200103931 modified "2023-09-27" @default.
- W3200103931 title "Efficient knowledge distillation of teacher model to multiple student models" @default.
- W3200103931 cites W1724438581 @default.
- W3200103931 cites W1821462560 @default.
- W3200103931 cites W2194775991 @default.
- W3200103931 cites W2294370754 @default.
- W3200103931 cites W2473930607 @default.
- W3200103931 cites W2508418541 @default.
- W3200103931 cites W2732039841 @default.
- W3200103931 cites W2808847742 @default.
- W3200103931 cites W2886014761 @default.
- W3200103931 cites W2904206140 @default.
- W3200103931 cites W2911803042 @default.
- W3200103931 cites W2924902521 @default.
- W3200103931 cites W2954886595 @default.
- W3200103931 cites W2963140444 @default.
- W3200103931 cites W2963387524 @default.
- W3200103931 cites W2964118293 @default.
- W3200103931 cites W2964121744 @default.
- W3200103931 cites W2969964734 @default.
- W3200103931 cites W2970454332 @default.
- W3200103931 cites W2970971581 @default.
- W3200103931 cites W2986015886 @default.
- W3200103931 cites W2991040477 @default.
- W3200103931 cites W2996999409 @default.
- W3200103931 cites W2997128429 @default.
- W3200103931 cites W2997224071 @default.
- W3200103931 cites W3033902397 @default.
- W3200103931 cites W3034406766 @default.
- W3200103931 cites W3104737411 @default.
- W3200103931 cites W3105966348 @default.
- W3200103931 cites W3106012737 @default.
- W3200103931 cites W3137609883 @default.
- W3200103931 cites W3174088783 @default.
- W3200103931 doi "https://doi.org/10.1109/iaict52856.2021.9532543" @default.
- W3200103931 hasPublicationYear "2021" @default.
- W3200103931 type Work @default.
- W3200103931 sameAs 3200103931 @default.
- W3200103931 citedByCount "0" @default.
- W3200103931 crossrefType "proceedings-article" @default.
- W3200103931 hasAuthorship W3200103931A5019643071 @default.
- W3200103931 hasAuthorship W3200103931A5023878418 @default.
- W3200103931 hasAuthorship W3200103931A5041130310 @default.
- W3200103931 hasAuthorship W3200103931A5060733879 @default.
- W3200103931 hasConcept C105339364 @default.
- W3200103931 hasConcept C115903868 @default.
- W3200103931 hasConcept C119857082 @default.
- W3200103931 hasConcept C14036430 @default.
- W3200103931 hasConcept C154945302 @default.
- W3200103931 hasConcept C162307627 @default.
- W3200103931 hasConcept C162324750 @default.
- W3200103931 hasConcept C177264268 @default.
- W3200103931 hasConcept C178790620 @default.
- W3200103931 hasConcept C185592680 @default.
- W3200103931 hasConcept C187736073 @default.
- W3200103931 hasConcept C199360897 @default.
- W3200103931 hasConcept C204030448 @default.
- W3200103931 hasConcept C2780451532 @default.
- W3200103931 hasConcept C41008148 @default.
- W3200103931 hasConcept C78458016 @default.
- W3200103931 hasConcept C86803240 @default.
- W3200103931 hasConceptScore W3200103931C105339364 @default.
- W3200103931 hasConceptScore W3200103931C115903868 @default.
- W3200103931 hasConceptScore W3200103931C119857082 @default.
- W3200103931 hasConceptScore W3200103931C14036430 @default.
- W3200103931 hasConceptScore W3200103931C154945302 @default.
- W3200103931 hasConceptScore W3200103931C162307627 @default.
- W3200103931 hasConceptScore W3200103931C162324750 @default.
- W3200103931 hasConceptScore W3200103931C177264268 @default.
- W3200103931 hasConceptScore W3200103931C178790620 @default.
- W3200103931 hasConceptScore W3200103931C185592680 @default.
- W3200103931 hasConceptScore W3200103931C187736073 @default.
- W3200103931 hasConceptScore W3200103931C199360897 @default.
- W3200103931 hasConceptScore W3200103931C204030448 @default.
- W3200103931 hasConceptScore W3200103931C2780451532 @default.
- W3200103931 hasConceptScore W3200103931C41008148 @default.
- W3200103931 hasConceptScore W3200103931C78458016 @default.
- W3200103931 hasConceptScore W3200103931C86803240 @default.
- W3200103931 hasLocation W32001039311 @default.
- W3200103931 hasOpenAccess W3200103931 @default.
- W3200103931 hasPrimaryLocation W32001039311 @default.
- W3200103931 hasRelatedWork W1969481115 @default.
- W3200103931 hasRelatedWork W2021850411 @default.
- W3200103931 hasRelatedWork W2072834482 @default.
- W3200103931 hasRelatedWork W2096848550 @default.
- W3200103931 hasRelatedWork W2961085424 @default.
- W3200103931 hasRelatedWork W4286629047 @default.
- W3200103931 hasRelatedWork W4306321456 @default.
- W3200103931 hasRelatedWork W4306674287 @default.
- W3200103931 hasRelatedWork W4312263439 @default.
- W3200103931 hasRelatedWork W4224009465 @default.
- W3200103931 isParatext "false" @default.
- W3200103931 isRetracted "false" @default.