SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3100156920> ?p ?o ?g. }

Showing items 1 to 83 of 83 with 100 items per page.

W3100156920 abstract "Multi-Class Incremental Learning (MCIL) aims to learn new concepts by incrementally updating a model trained on previous concepts. However, there is an inherent trade-off to effectively learning new concepts without catastrophic forgetting of previous ones. To alleviate this issue, it has been proposed to keep around a few examples of the previous concepts but the effectiveness of this approach heavily depends on the representativeness of these examples. This paper proposes a novel and automatic framework we call mnemonics, where we parameterize exemplars and make them optimizable in an end-to-end manner. We train the framework through bilevel optimizations, i.e., model-level and exemplar-level. We conduct extensive experiments on three MCIL benchmarks, CIFAR-100, ImageNet-Subset and ImageNet, and show that using mnemonics exemplars can surpass the state-of-the-art by a large margin. Interestingly and quite intriguingly, the mnemonics exemplars tend to be on the boundaries between different classes." @default.
W3100156920 created "2020-11-23" @default.
W3100156920 creator A5033713097 @default.
W3100156920 creator A5034434932 @default.
W3100156920 creator A5051534545 @default.
W3100156920 creator A5081485810 @default.
W3100156920 creator A5083321698 @default.
W3100156920 date "2020-06-01" @default.
W3100156920 modified "2023-10-16" @default.
W3100156920 title "Mnemonics Training: Multi-Class Incremental Learning Without Forgetting" @default.
W3100156920 cites W1682403713 @default.
W3100156920 cites W1997865285 @default.
W3100156920 cites W2015563892 @default.
W3100156920 cites W2117539524 @default.
W3100156920 cites W2166344886 @default.
W3100156920 cites W2194775991 @default.
W3100156920 cites W2473930607 @default.
W3100156920 cites W2614367549 @default.
W3100156920 cites W2948734064 @default.
W3100156920 cites W2954929116 @default.
W3100156920 cites W2963599420 @default.
W3100156920 cites W2963943197 @default.
W3100156920 cites W2964189064 @default.
W3100156920 doi "https://doi.org/10.1109/cvpr42600.2020.01226" @default.
W3100156920 hasPublicationYear "2020" @default.
W3100156920 type Work @default.
W3100156920 sameAs 3100156920 @default.
W3100156920 citedByCount "154" @default.
W3100156920 countsByYear W31001569202019 @default.
W3100156920 countsByYear W31001569202020 @default.
W3100156920 countsByYear W31001569202021 @default.
W3100156920 countsByYear W31001569202022 @default.
W3100156920 countsByYear W31001569202023 @default.
W3100156920 crossrefType "proceedings-article" @default.
W3100156920 hasAuthorship W3100156920A5033713097 @default.
W3100156920 hasAuthorship W3100156920A5034434932 @default.
W3100156920 hasAuthorship W3100156920A5051534545 @default.
W3100156920 hasAuthorship W3100156920A5081485810 @default.
W3100156920 hasAuthorship W3100156920A5083321698 @default.
W3100156920 hasBestOaLocation W31001569202 @default.
W3100156920 hasConcept C119857082 @default.
W3100156920 hasConcept C154945302 @default.
W3100156920 hasConcept C15744967 @default.
W3100156920 hasConcept C180747234 @default.
W3100156920 hasConcept C197792726 @default.
W3100156920 hasConcept C2777212361 @default.
W3100156920 hasConcept C37381756 @default.
W3100156920 hasConcept C41008148 @default.
W3100156920 hasConcept C7149132 @default.
W3100156920 hasConcept C774472 @default.
W3100156920 hasConcept C77805123 @default.
W3100156920 hasConcept C97931131 @default.
W3100156920 hasConceptScore W3100156920C119857082 @default.
W3100156920 hasConceptScore W3100156920C154945302 @default.
W3100156920 hasConceptScore W3100156920C15744967 @default.
W3100156920 hasConceptScore W3100156920C180747234 @default.
W3100156920 hasConceptScore W3100156920C197792726 @default.
W3100156920 hasConceptScore W3100156920C2777212361 @default.
W3100156920 hasConceptScore W3100156920C37381756 @default.
W3100156920 hasConceptScore W3100156920C41008148 @default.
W3100156920 hasConceptScore W3100156920C7149132 @default.
W3100156920 hasConceptScore W3100156920C774472 @default.
W3100156920 hasConceptScore W3100156920C77805123 @default.
W3100156920 hasConceptScore W3100156920C97931131 @default.
W3100156920 hasLocation W31001569201 @default.
W3100156920 hasLocation W31001569202 @default.
W3100156920 hasLocation W31001569203 @default.
W3100156920 hasOpenAccess W3100156920 @default.
W3100156920 hasPrimaryLocation W31001569201 @default.
W3100156920 hasRelatedWork W2095758845 @default.
W3100156920 hasRelatedWork W2731757110 @default.
W3100156920 hasRelatedWork W2934987099 @default.
W3100156920 hasRelatedWork W3014026011 @default.
W3100156920 hasRelatedWork W3100156920 @default.
W3100156920 hasRelatedWork W3133105109 @default.
W3100156920 hasRelatedWork W3174101157 @default.
W3100156920 hasRelatedWork W4287326170 @default.
W3100156920 hasRelatedWork W4287755010 @default.
W3100156920 hasRelatedWork W4319994054 @default.
W3100156920 isParatext "false" @default.
W3100156920 isRetracted "false" @default.
W3100156920 magId "3100156920" @default.
W3100156920 workType "article" @default.