Matches in SemOpenAlex for { <https://semopenalex.org/work/W2890619523> ?p ?o ?g. }
- W2890619523 abstract "Metric learning aims at learning a distance which is consistent with the semantic meaning of the samples. The problem is generally solved by learning an embedding for each sample such that the embeddings of samples of the same category are compact while the embeddings of samples of different categories are spread-out in the feature space. We study the features extracted from the second last layer of a deep neural network based classifier trained with the cross entropy loss on top of the softmax layer. We show that training classifiers with different temperature values of softmax function leads to features with different levels of compactness. Leveraging these insights, we propose a heating-up strategy to train a classifier with increasing temperatures, leading the corresponding embeddings to achieve state-of-the-art performance on a variety of metric learning benchmarks." @default.
- W2890619523 created "2018-09-27" @default.
- W2890619523 creator A5008582408 @default.
- W2890619523 creator A5050615258 @default.
- W2890619523 creator A5065795593 @default.
- W2890619523 creator A5067090814 @default.
- W2890619523 creator A5073568638 @default.
- W2890619523 date "2018-09-27" @default.
- W2890619523 modified "2023-09-22" @default.
- W2890619523 title "Heated-Up Softmax Embedding." @default.
- W2890619523 cites W1532325895 @default.
- W2890619523 cites W1821462560 @default.
- W2890619523 cites W2068042582 @default.
- W2890619523 cites W2117154949 @default.
- W2890619523 cites W2117539524 @default.
- W2890619523 cites W2138011018 @default.
- W2890619523 cites W2148214025 @default.
- W2890619523 cites W2151483828 @default.
- W2890619523 cites W2547446130 @default.
- W2890619523 cites W2549858646 @default.
- W2890619523 cites W2749130292 @default.
- W2890619523 cites W2949117887 @default.
- W2890619523 cites W2950188904 @default.
- W2890619523 cites W2953391683 @default.
- W2890619523 cites W2963026686 @default.
- W2890619523 cites W2963988212 @default.
- W2890619523 cites W2600537992 @default.
- W2890619523 hasPublicationYear "2018" @default.
- W2890619523 type Work @default.
- W2890619523 sameAs 2890619523 @default.
- W2890619523 citedByCount "19" @default.
- W2890619523 countsByYear W28906195232019 @default.
- W2890619523 countsByYear W28906195232020 @default.
- W2890619523 countsByYear W28906195232021 @default.
- W2890619523 crossrefType "posted-content" @default.
- W2890619523 hasAuthorship W2890619523A5008582408 @default.
- W2890619523 hasAuthorship W2890619523A5050615258 @default.
- W2890619523 hasAuthorship W2890619523A5065795593 @default.
- W2890619523 hasAuthorship W2890619523A5067090814 @default.
- W2890619523 hasAuthorship W2890619523A5073568638 @default.
- W2890619523 hasConcept C118615104 @default.
- W2890619523 hasConcept C119857082 @default.
- W2890619523 hasConcept C127413603 @default.
- W2890619523 hasConcept C134306372 @default.
- W2890619523 hasConcept C153180895 @default.
- W2890619523 hasConcept C154945302 @default.
- W2890619523 hasConcept C176217482 @default.
- W2890619523 hasConcept C18648836 @default.
- W2890619523 hasConcept C188441871 @default.
- W2890619523 hasConcept C198043062 @default.
- W2890619523 hasConcept C21547014 @default.
- W2890619523 hasConcept C33923547 @default.
- W2890619523 hasConcept C41008148 @default.
- W2890619523 hasConcept C41608201 @default.
- W2890619523 hasConcept C50644808 @default.
- W2890619523 hasConcept C83665646 @default.
- W2890619523 hasConcept C95623464 @default.
- W2890619523 hasConceptScore W2890619523C118615104 @default.
- W2890619523 hasConceptScore W2890619523C119857082 @default.
- W2890619523 hasConceptScore W2890619523C127413603 @default.
- W2890619523 hasConceptScore W2890619523C134306372 @default.
- W2890619523 hasConceptScore W2890619523C153180895 @default.
- W2890619523 hasConceptScore W2890619523C154945302 @default.
- W2890619523 hasConceptScore W2890619523C176217482 @default.
- W2890619523 hasConceptScore W2890619523C18648836 @default.
- W2890619523 hasConceptScore W2890619523C188441871 @default.
- W2890619523 hasConceptScore W2890619523C198043062 @default.
- W2890619523 hasConceptScore W2890619523C21547014 @default.
- W2890619523 hasConceptScore W2890619523C33923547 @default.
- W2890619523 hasConceptScore W2890619523C41008148 @default.
- W2890619523 hasConceptScore W2890619523C41608201 @default.
- W2890619523 hasConceptScore W2890619523C50644808 @default.
- W2890619523 hasConceptScore W2890619523C83665646 @default.
- W2890619523 hasConceptScore W2890619523C95623464 @default.
- W2890619523 hasLocation W28906195231 @default.
- W2890619523 hasOpenAccess W2890619523 @default.
- W2890619523 hasPrimaryLocation W28906195231 @default.
- W2890619523 hasRelatedWork W1821462560 @default.
- W2890619523 hasRelatedWork W1836465849 @default.
- W2890619523 hasRelatedWork W2108598243 @default.
- W2890619523 hasRelatedWork W2117539524 @default.
- W2890619523 hasRelatedWork W2138011018 @default.
- W2890619523 hasRelatedWork W2138621090 @default.
- W2890619523 hasRelatedWork W2194775991 @default.
- W2890619523 hasRelatedWork W2555897561 @default.
- W2890619523 hasRelatedWork W2598634450 @default.
- W2890619523 hasRelatedWork W2810967842 @default.
- W2890619523 hasRelatedWork W2895347732 @default.
- W2890619523 hasRelatedWork W2899771611 @default.
- W2890619523 hasRelatedWork W2933502442 @default.
- W2890619523 hasRelatedWork W2952494021 @default.
- W2890619523 hasRelatedWork W2963026686 @default.
- W2890619523 hasRelatedWork W2963350250 @default.
- W2890619523 hasRelatedWork W2963466847 @default.
- W2890619523 hasRelatedWork W3099206234 @default.
- W2890619523 hasRelatedWork W3100506510 @default.
- W2890619523 hasRelatedWork W2600537992 @default.
- W2890619523 isParatext "false" @default.
- W2890619523 isRetracted "false" @default.
- W2890619523 magId "2890619523" @default.