Matches in SemOpenAlex for { <https://semopenalex.org/work/W4221016573> ?p ?o ?g. }
- W4221016573 endingPage "16650" @default.
- W4221016573 startingPage "16640" @default.
- W4221016573 abstract "Data augmentation has been proved effective in training deep models. Existing data augmentation methods tackle the fine-grained problem by blending image pairs and fusing corresponding labels according to the statistics of mixed pixels, which produces additional noise harmful to the performance of networks. Motivated by this, we present a simple yet effective cross ensemble knowledge distillation (CEKD) model for fine-grained feature learning. We innovatively propose a cross distillation module to provide additional supervision to alleviate the noise problem, and propose a collaborative ensemble module to overcome the target conflict problem. The proposed model can be trained in an end-to-end manner, and only requires image-level label supervision. Extensive experiments on widely used fine-grained benchmarks demonstrate the effectiveness of our proposed model. Specifically, with the backbone of ResNet-101, CEKD obtains the accuracy of 89.59%, 95.96% and 94.56% in three datasets respectively, outperforming state-of-the-art API-Net by 0.99%, 1.06% and 1.16%." @default.
- W4221016573 created "2022-04-03" @default.
- W4221016573 creator A5005075206 @default.
- W4221016573 creator A5017818991 @default.
- W4221016573 creator A5030064573 @default.
- W4221016573 creator A5042431015 @default.
- W4221016573 creator A5047303769 @default.
- W4221016573 creator A5052124467 @default.
- W4221016573 date "2022-03-24" @default.
- W4221016573 modified "2023-10-18" @default.
- W4221016573 title "CEKD:Cross ensemble knowledge distillation for augmented fine-grained data" @default.
- W4221016573 cites W1955942245 @default.
- W4221016573 cites W1994488211 @default.
- W4221016573 cites W2138011018 @default.
- W4221016573 cites W2183341477 @default.
- W4221016573 cites W2194775991 @default.
- W4221016573 cites W2202499615 @default.
- W4221016573 cites W2295107390 @default.
- W4221016573 cites W2737725206 @default.
- W4221016573 cites W2752782242 @default.
- W4221016573 cites W2773003563 @default.
- W4221016573 cites W2807931652 @default.
- W4221016573 cites W2887783173 @default.
- W4221016573 cites W2891951760 @default.
- W4221016573 cites W2961018736 @default.
- W4221016573 cites W2962858109 @default.
- W4221016573 cites W2963446712 @default.
- W4221016573 cites W2963622428 @default.
- W4221016573 cites W2964016283 @default.
- W4221016573 cites W2964274719 @default.
- W4221016573 cites W2986821660 @default.
- W4221016573 cites W2992308087 @default.
- W4221016573 cites W2996970889 @default.
- W4221016573 cites W2997006708 @default.
- W4221016573 cites W2997426000 @default.
- W4221016573 cites W2998619563 @default.
- W4221016573 cites W3009073662 @default.
- W4221016573 cites W3033250778 @default.
- W4221016573 cites W3034342078 @default.
- W4221016573 cites W3034865111 @default.
- W4221016573 cites W3035321581 @default.
- W4221016573 cites W3043613541 @default.
- W4221016573 cites W3099920683 @default.
- W4221016573 cites W3108870912 @default.
- W4221016573 cites W3128325346 @default.
- W4221016573 cites W3175248300 @default.
- W4221016573 cites W3177330184 @default.
- W4221016573 cites W4234674466 @default.
- W4221016573 cites W56385144 @default.
- W4221016573 doi "https://doi.org/10.1007/s10489-022-03355-0" @default.
- W4221016573 hasPublicationYear "2022" @default.
- W4221016573 type Work @default.
- W4221016573 citedByCount "0" @default.
- W4221016573 crossrefType "journal-article" @default.
- W4221016573 hasAuthorship W4221016573A5005075206 @default.
- W4221016573 hasAuthorship W4221016573A5017818991 @default.
- W4221016573 hasAuthorship W4221016573A5030064573 @default.
- W4221016573 hasAuthorship W4221016573A5042431015 @default.
- W4221016573 hasAuthorship W4221016573A5047303769 @default.
- W4221016573 hasAuthorship W4221016573A5052124467 @default.
- W4221016573 hasBestOaLocation W42210165732 @default.
- W4221016573 hasConcept C108583219 @default.
- W4221016573 hasConcept C111472728 @default.
- W4221016573 hasConcept C115961682 @default.
- W4221016573 hasConcept C119857082 @default.
- W4221016573 hasConcept C124101348 @default.
- W4221016573 hasConcept C138885662 @default.
- W4221016573 hasConcept C153180895 @default.
- W4221016573 hasConcept C154945302 @default.
- W4221016573 hasConcept C160633673 @default.
- W4221016573 hasConcept C178790620 @default.
- W4221016573 hasConcept C185592680 @default.
- W4221016573 hasConcept C204030448 @default.
- W4221016573 hasConcept C2776401178 @default.
- W4221016573 hasConcept C2780586882 @default.
- W4221016573 hasConcept C41008148 @default.
- W4221016573 hasConcept C41895202 @default.
- W4221016573 hasConcept C45942800 @default.
- W4221016573 hasConcept C99498987 @default.
- W4221016573 hasConceptScore W4221016573C108583219 @default.
- W4221016573 hasConceptScore W4221016573C111472728 @default.
- W4221016573 hasConceptScore W4221016573C115961682 @default.
- W4221016573 hasConceptScore W4221016573C119857082 @default.
- W4221016573 hasConceptScore W4221016573C124101348 @default.
- W4221016573 hasConceptScore W4221016573C138885662 @default.
- W4221016573 hasConceptScore W4221016573C153180895 @default.
- W4221016573 hasConceptScore W4221016573C154945302 @default.
- W4221016573 hasConceptScore W4221016573C160633673 @default.
- W4221016573 hasConceptScore W4221016573C178790620 @default.
- W4221016573 hasConceptScore W4221016573C185592680 @default.
- W4221016573 hasConceptScore W4221016573C204030448 @default.
- W4221016573 hasConceptScore W4221016573C2776401178 @default.
- W4221016573 hasConceptScore W4221016573C2780586882 @default.
- W4221016573 hasConceptScore W4221016573C41008148 @default.
- W4221016573 hasConceptScore W4221016573C41895202 @default.
- W4221016573 hasConceptScore W4221016573C45942800 @default.
- W4221016573 hasConceptScore W4221016573C99498987 @default.
- W4221016573 hasFunder F4320321001 @default.