Matches in SemOpenAlex for { <https://semopenalex.org/work/W4310581307> ?p ?o ?g. }
Showing items 1 to 72 of
72
with 100 items per page.
- W4310581307 abstract "Deep neural network has a significant improvement in accuracy for object detection. A great challenge for deep learning is deployment on low-latency inference systems. Knowledge distillation is widely used for the reduced computational complexity and the compatibility with embedded hardware devices. In this paper, we proposal two adaptive balance distillation methods for object detection, Positive-Feature Balance Distillation and Hard-Feature Balance Distillation. Positive-Feature Balance Distillation helps to alleviate the imbalance of the positive and negative feature and Hard-Feature Balance Distillation is used for forcing the student to focus on the tiny amounts of hard feature in object detection task. These methods are used to improve the accuracy of student networks by transferring knowledge from a complicated teacher’s network to a simplified student’s network. We conduct comprehensive empirical evaluation with different knowledge distillation configurations over KITTI that is one of the most important automatic driving datasets. The proposed methods outperform consistent improvement in accuracy-speed trade-offs for modern object detection method." @default.
- W4310581307 created "2022-12-12" @default.
- W4310581307 creator A5033743267 @default.
- W4310581307 creator A5040009629 @default.
- W4310581307 creator A5077042443 @default.
- W4310581307 date "2022-10-21" @default.
- W4310581307 modified "2023-09-27" @default.
- W4310581307 title "Distilling the Knowledge in Object Detection with Adaptive Balance" @default.
- W4310581307 cites W2115579991 @default.
- W4310581307 cites W2194775991 @default.
- W4310581307 cites W2750432752 @default.
- W4310581307 cites W2959289524 @default.
- W4310581307 cites W2962914239 @default.
- W4310581307 cites W2963323244 @default.
- W4310581307 cites W2963351448 @default.
- W4310581307 cites W3034368386 @default.
- W4310581307 cites W3173270634 @default.
- W4310581307 cites W3176459575 @default.
- W4310581307 cites W3201877444 @default.
- W4310581307 cites W4313141028 @default.
- W4310581307 doi "https://doi.org/10.1109/icsp56322.2022.9965214" @default.
- W4310581307 hasPublicationYear "2022" @default.
- W4310581307 type Work @default.
- W4310581307 citedByCount "0" @default.
- W4310581307 crossrefType "proceedings-article" @default.
- W4310581307 hasAuthorship W4310581307A5033743267 @default.
- W4310581307 hasAuthorship W4310581307A5040009629 @default.
- W4310581307 hasAuthorship W4310581307A5077042443 @default.
- W4310581307 hasConcept C119857082 @default.
- W4310581307 hasConcept C138885662 @default.
- W4310581307 hasConcept C153180895 @default.
- W4310581307 hasConcept C154945302 @default.
- W4310581307 hasConcept C178790620 @default.
- W4310581307 hasConcept C185592680 @default.
- W4310581307 hasConcept C204030448 @default.
- W4310581307 hasConcept C2776151529 @default.
- W4310581307 hasConcept C2776214188 @default.
- W4310581307 hasConcept C2776401178 @default.
- W4310581307 hasConcept C41008148 @default.
- W4310581307 hasConcept C41895202 @default.
- W4310581307 hasConcept C50644808 @default.
- W4310581307 hasConcept C52622490 @default.
- W4310581307 hasConceptScore W4310581307C119857082 @default.
- W4310581307 hasConceptScore W4310581307C138885662 @default.
- W4310581307 hasConceptScore W4310581307C153180895 @default.
- W4310581307 hasConceptScore W4310581307C154945302 @default.
- W4310581307 hasConceptScore W4310581307C178790620 @default.
- W4310581307 hasConceptScore W4310581307C185592680 @default.
- W4310581307 hasConceptScore W4310581307C204030448 @default.
- W4310581307 hasConceptScore W4310581307C2776151529 @default.
- W4310581307 hasConceptScore W4310581307C2776214188 @default.
- W4310581307 hasConceptScore W4310581307C2776401178 @default.
- W4310581307 hasConceptScore W4310581307C41008148 @default.
- W4310581307 hasConceptScore W4310581307C41895202 @default.
- W4310581307 hasConceptScore W4310581307C50644808 @default.
- W4310581307 hasConceptScore W4310581307C52622490 @default.
- W4310581307 hasLocation W43105813071 @default.
- W4310581307 hasOpenAccess W4310581307 @default.
- W4310581307 hasPrimaryLocation W43105813071 @default.
- W4310581307 hasRelatedWork W2016461833 @default.
- W4310581307 hasRelatedWork W2059299633 @default.
- W4310581307 hasRelatedWork W2132943804 @default.
- W4310581307 hasRelatedWork W2144059113 @default.
- W4310581307 hasRelatedWork W2146076056 @default.
- W4310581307 hasRelatedWork W2382607599 @default.
- W4310581307 hasRelatedWork W2811390910 @default.
- W4310581307 hasRelatedWork W3003836766 @default.
- W4310581307 hasRelatedWork W3197541072 @default.
- W4310581307 hasRelatedWork W4295357754 @default.
- W4310581307 isParatext "false" @default.
- W4310581307 isRetracted "false" @default.
- W4310581307 workType "article" @default.