Matches in SemOpenAlex for { <https://semopenalex.org/work/W2894568782> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W2894568782 endingPage "86" @default.
- W2894568782 startingPage "71" @default.
- W2894568782 abstract "Triplet loss, popular for metric learning, has made a great success in many computer vision tasks, such as fine-grained image classification, image retrieval, and face recognition. Considering that the number of triplets grows cubically with the size of training data, triplet selection is thus indispensable for efficiently training with triplet loss. However, in practice, the training is usually very sensitive to the selection of triplets, e.g., it almost does not converge with randomly selected triplets and selecting the hardest triplets also leads to bad local minima. We argue that the bias in the selection of triplets degrades the performance of learning with triplet loss. In this paper, we propose a new variant of triplet loss, which tries to reduce the bias in triplet selection by adaptively correcting the distribution shift on the selected triplets. We refer to this new triplet loss as adapted triplet loss. We conduct a number of experiments on MNIST and Fashion-MNIST for image classification, and on CARS196, CUB200-2011, and Stanford Online Products for image retrieval. The experimental results demonstrate the effectiveness of the proposed method." @default.
- W2894568782 created "2018-10-12" @default.
- W2894568782 creator A5038748720 @default.
- W2894568782 creator A5065250332 @default.
- W2894568782 creator A5074103823 @default.
- W2894568782 creator A5085309099 @default.
- W2894568782 creator A5091227928 @default.
- W2894568782 date "2018-01-01" @default.
- W2894568782 modified "2023-09-30" @default.
- W2894568782 title "Correcting the Triplet Selection Bias for Triplet Loss" @default.
- W2894568782 cites W1677182931 @default.
- W2894568782 cites W1892016050 @default.
- W2894568782 cites W1939575207 @default.
- W2894568782 cites W1971955426 @default.
- W2894568782 cites W1975517671 @default.
- W2894568782 cites W2068074736 @default.
- W2894568782 cites W2097117768 @default.
- W2894568782 cites W2104094955 @default.
- W2894568782 cites W2108598243 @default.
- W2894568782 cites W2112796928 @default.
- W2894568782 cites W2115403315 @default.
- W2894568782 cites W2138011018 @default.
- W2894568782 cites W2155893237 @default.
- W2894568782 cites W2170881581 @default.
- W2894568782 cites W2325939864 @default.
- W2894568782 cites W2340690086 @default.
- W2894568782 cites W2342611082 @default.
- W2894568782 cites W2467139031 @default.
- W2894568782 cites W2470322391 @default.
- W2894568782 cites W2480418144 @default.
- W2894568782 cites W2519373641 @default.
- W2894568782 cites W2549858646 @default.
- W2894568782 cites W2606377603 @default.
- W2894568782 cites W2740168373 @default.
- W2894568782 cites W2951019013 @default.
- W2894568782 cites W2962798895 @default.
- W2894568782 cites W2963026686 @default.
- W2894568782 cites W2963350250 @default.
- W2894568782 cites W2963389687 @default.
- W2894568782 cites W2963775347 @default.
- W2894568782 cites W2964076257 @default.
- W2894568782 cites W2964124968 @default.
- W2894568782 cites W3099206234 @default.
- W2894568782 cites W3100555577 @default.
- W2894568782 cites W3103919331 @default.
- W2894568782 doi "https://doi.org/10.1007/978-3-030-01231-1_5" @default.
- W2894568782 hasPublicationYear "2018" @default.
- W2894568782 type Work @default.
- W2894568782 sameAs 2894568782 @default.
- W2894568782 citedByCount "58" @default.
- W2894568782 countsByYear W28945687822019 @default.
- W2894568782 countsByYear W28945687822020 @default.
- W2894568782 countsByYear W28945687822021 @default.
- W2894568782 countsByYear W28945687822022 @default.
- W2894568782 countsByYear W28945687822023 @default.
- W2894568782 crossrefType "book-chapter" @default.
- W2894568782 hasAuthorship W2894568782A5038748720 @default.
- W2894568782 hasAuthorship W2894568782A5065250332 @default.
- W2894568782 hasAuthorship W2894568782A5074103823 @default.
- W2894568782 hasAuthorship W2894568782A5085309099 @default.
- W2894568782 hasAuthorship W2894568782A5091227928 @default.
- W2894568782 hasConcept C154945302 @default.
- W2894568782 hasConcept C41008148 @default.
- W2894568782 hasConcept C81917197 @default.
- W2894568782 hasConceptScore W2894568782C154945302 @default.
- W2894568782 hasConceptScore W2894568782C41008148 @default.
- W2894568782 hasConceptScore W2894568782C81917197 @default.
- W2894568782 hasLocation W28945687821 @default.
- W2894568782 hasOpenAccess W2894568782 @default.
- W2894568782 hasPrimaryLocation W28945687821 @default.
- W2894568782 hasRelatedWork W1485063768 @default.
- W2894568782 hasRelatedWork W1990422213 @default.
- W2894568782 hasRelatedWork W2115461783 @default.
- W2894568782 hasRelatedWork W2118336583 @default.
- W2894568782 hasRelatedWork W2358668433 @default.
- W2894568782 hasRelatedWork W2390279801 @default.
- W2894568782 hasRelatedWork W2748952813 @default.
- W2894568782 hasRelatedWork W2899084033 @default.
- W2894568782 hasRelatedWork W2970622411 @default.
- W2894568782 hasRelatedWork W2751504412 @default.
- W2894568782 isParatext "false" @default.
- W2894568782 isRetracted "false" @default.
- W2894568782 magId "2894568782" @default.
- W2894568782 workType "book-chapter" @default.