Matches in SemOpenAlex for { <https://semopenalex.org/work/W4283803536> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W4283803536 endingPage "12936" @default.
- W4283803536 startingPage "12935" @default.
- W4283803536 abstract "Capturing visual similarity among images is the core of many computer vision and pattern recognition tasks. This problem can be formulated in such a paradigm called metric learning. Most research in the area has been mainly focusing on improving the loss functions and similarity measures. However, due to the ignoring of geometric structure, existing methods often lead to sub-optimal results. Thus, several recent research methods took advantage of Wasserstein distance between batches of samples to characterize the spacial geometry. Although these approaches can achieve enhanced performance, the aggregation over batches definitely hinders Wasserstein distance's superior measure capability and leads to high computational complexity. To address this limitation, we propose a novel Deep Wasserstein Metric Learning framework, which employs Wasserstein distance to precisely capture the relationship among various images under ranking-based loss functions such as contrastive loss and triplet loss. Our method directly computes the distance between images, considering the geometry at a finer granularity than batch level. Furthermore, we introduce a new efficient algorithm using Sinkhorn approximation and Wasserstein measure coreset. The experimental results demonstrate the improvements of our framework over various baselines in different applications and benchmark datasets." @default.
- W4283803536 created "2022-07-05" @default.
- W4283803536 creator A5009311791 @default.
- W4283803536 creator A5012652927 @default.
- W4283803536 creator A5084946861 @default.
- W4283803536 date "2022-06-28" @default.
- W4283803536 modified "2023-09-30" @default.
- W4283803536 title "An Optimal Transport Approach to Deep Metric Learning (Student Abstract)" @default.
- W4283803536 doi "https://doi.org/10.1609/aaai.v36i11.21604" @default.
- W4283803536 hasPublicationYear "2022" @default.
- W4283803536 type Work @default.
- W4283803536 citedByCount "3" @default.
- W4283803536 countsByYear W42838035362023 @default.
- W4283803536 crossrefType "journal-article" @default.
- W4283803536 hasAuthorship W4283803536A5009311791 @default.
- W4283803536 hasAuthorship W4283803536A5012652927 @default.
- W4283803536 hasAuthorship W4283803536A5084946861 @default.
- W4283803536 hasBestOaLocation W42838035361 @default.
- W4283803536 hasConcept C103278499 @default.
- W4283803536 hasConcept C108583219 @default.
- W4283803536 hasConcept C111919701 @default.
- W4283803536 hasConcept C11413529 @default.
- W4283803536 hasConcept C115961682 @default.
- W4283803536 hasConcept C119857082 @default.
- W4283803536 hasConcept C124101348 @default.
- W4283803536 hasConcept C126255220 @default.
- W4283803536 hasConcept C13280743 @default.
- W4283803536 hasConcept C154945302 @default.
- W4283803536 hasConcept C162324750 @default.
- W4283803536 hasConcept C176217482 @default.
- W4283803536 hasConcept C177774035 @default.
- W4283803536 hasConcept C185798385 @default.
- W4283803536 hasConcept C189430467 @default.
- W4283803536 hasConcept C205649164 @default.
- W4283803536 hasConcept C21547014 @default.
- W4283803536 hasConcept C2776517306 @default.
- W4283803536 hasConcept C2777634741 @default.
- W4283803536 hasConcept C2780009758 @default.
- W4283803536 hasConcept C28826006 @default.
- W4283803536 hasConcept C33923547 @default.
- W4283803536 hasConcept C41008148 @default.
- W4283803536 hasConcept C80444323 @default.
- W4283803536 hasConcept C82668687 @default.
- W4283803536 hasConceptScore W4283803536C103278499 @default.
- W4283803536 hasConceptScore W4283803536C108583219 @default.
- W4283803536 hasConceptScore W4283803536C111919701 @default.
- W4283803536 hasConceptScore W4283803536C11413529 @default.
- W4283803536 hasConceptScore W4283803536C115961682 @default.
- W4283803536 hasConceptScore W4283803536C119857082 @default.
- W4283803536 hasConceptScore W4283803536C124101348 @default.
- W4283803536 hasConceptScore W4283803536C126255220 @default.
- W4283803536 hasConceptScore W4283803536C13280743 @default.
- W4283803536 hasConceptScore W4283803536C154945302 @default.
- W4283803536 hasConceptScore W4283803536C162324750 @default.
- W4283803536 hasConceptScore W4283803536C176217482 @default.
- W4283803536 hasConceptScore W4283803536C177774035 @default.
- W4283803536 hasConceptScore W4283803536C185798385 @default.
- W4283803536 hasConceptScore W4283803536C189430467 @default.
- W4283803536 hasConceptScore W4283803536C205649164 @default.
- W4283803536 hasConceptScore W4283803536C21547014 @default.
- W4283803536 hasConceptScore W4283803536C2776517306 @default.
- W4283803536 hasConceptScore W4283803536C2777634741 @default.
- W4283803536 hasConceptScore W4283803536C2780009758 @default.
- W4283803536 hasConceptScore W4283803536C28826006 @default.
- W4283803536 hasConceptScore W4283803536C33923547 @default.
- W4283803536 hasConceptScore W4283803536C41008148 @default.
- W4283803536 hasConceptScore W4283803536C80444323 @default.
- W4283803536 hasConceptScore W4283803536C82668687 @default.
- W4283803536 hasIssue "11" @default.
- W4283803536 hasLocation W42838035361 @default.
- W4283803536 hasOpenAccess W4283803536 @default.
- W4283803536 hasPrimaryLocation W42838035361 @default.
- W4283803536 hasRelatedWork W2111173507 @default.
- W4283803536 hasRelatedWork W2319693127 @default.
- W4283803536 hasRelatedWork W2606618787 @default.
- W4283803536 hasRelatedWork W4223943233 @default.
- W4283803536 hasRelatedWork W4248020907 @default.
- W4283803536 hasRelatedWork W4283803536 @default.
- W4283803536 hasRelatedWork W4306742486 @default.
- W4283803536 hasRelatedWork W4312200629 @default.
- W4283803536 hasRelatedWork W4360585206 @default.
- W4283803536 hasRelatedWork W4380075502 @default.
- W4283803536 hasVolume "36" @default.
- W4283803536 isParatext "false" @default.
- W4283803536 isRetracted "false" @default.
- W4283803536 workType "article" @default.