Matches in SemOpenAlex for { <https://semopenalex.org/work/W2913505554> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W2913505554 abstract "Good quality similarity metrics can significantly facilitate the performance of many large-scale, real-world applications. Existing studies have proposed various solutions to learn a Mahalanobis or bilinear metric in an online fashion by either restricting distances between similar (dissimilar) pairs to be smaller (larger) than a given lower (upper) bound or requiring similar instances to be separated from dissimilar instances with a given margin. However, these linear metrics learned by leveraging fixed bounds or margins may not perform well in real-world applications, especially when data distributions are complex. We aim to address the open challenge of Online Adaptive Metric Learning (OAML) for learning adaptive metric functions on the fly. Unlike traditional online metric learning methods, OAML is significantly more challenging since the learned metric could be non-linear and the model has to be self-adaptive as more instances are observed. In this paper, we present a new online metric learning framework that attempts to tackle the challenge by learning an ANN-based metric with adaptive model complexity from a stream of constraints. In particular, we propose a novel Adaptive-Bound Triplet Loss (ABTL) to effectively utilize the input constraints and present a novel Adaptive Hedge Update (AHU) method for online updating the model parameters. We empirically validate the effectiveness and efficacy of our framework on various applications such as real-world image classification, facial verification, and image retrieval." @default.
- W2913505554 created "2019-02-21" @default.
- W2913505554 creator A5005002693 @default.
- W2913505554 creator A5013420855 @default.
- W2913505554 creator A5056179290 @default.
- W2913505554 creator A5067301238 @default.
- W2913505554 creator A5072193842 @default.
- W2913505554 date "2021-04-03" @default.
- W2913505554 modified "2023-09-24" @default.
- W2913505554 title "Towards Self-Adaptive Metric Learning On the Fly" @default.
- W2913505554 cites W1532499126 @default.
- W2913505554 cites W1898424075 @default.
- W2913505554 cites W1988790447 @default.
- W2913505554 cites W2003677307 @default.
- W2913505554 cites W2116731705 @default.
- W2913505554 cites W2117539524 @default.
- W2913505554 cites W2118979615 @default.
- W2913505554 cites W2124509324 @default.
- W2913505554 cites W2130556178 @default.
- W2913505554 cites W2138011018 @default.
- W2913505554 cites W2164408337 @default.
- W2913505554 cites W2170682041 @default.
- W2913505554 cites W2178031510 @default.
- W2913505554 cites W2331143823 @default.
- W2913505554 cites W2525990188 @default.
- W2913505554 cites W2536626143 @default.
- W2913505554 cites W2590796488 @default.
- W2913505554 cites W2750384547 @default.
- W2913505554 cites W2963055657 @default.
- W2913505554 cites W2963739929 @default.
- W2913505554 cites W3038013616 @default.
- W2913505554 cites W3118608800 @default.
- W2913505554 cites W3210232381 @default.
- W2913505554 doi "https://doi.org/10.48550/arxiv.2104.01495" @default.
- W2913505554 hasPublicationYear "2021" @default.
- W2913505554 type Work @default.
- W2913505554 sameAs 2913505554 @default.
- W2913505554 citedByCount "5" @default.
- W2913505554 countsByYear W29135055542018 @default.
- W2913505554 countsByYear W29135055542019 @default.
- W2913505554 countsByYear W29135055542020 @default.
- W2913505554 countsByYear W29135055542021 @default.
- W2913505554 crossrefType "posted-content" @default.
- W2913505554 hasAuthorship W2913505554A5005002693 @default.
- W2913505554 hasAuthorship W2913505554A5013420855 @default.
- W2913505554 hasAuthorship W2913505554A5056179290 @default.
- W2913505554 hasAuthorship W2913505554A5067301238 @default.
- W2913505554 hasAuthorship W2913505554A5072193842 @default.
- W2913505554 hasBestOaLocation W29135055541 @default.
- W2913505554 hasConcept C103278499 @default.
- W2913505554 hasConcept C115961682 @default.
- W2913505554 hasConcept C119857082 @default.
- W2913505554 hasConcept C134306372 @default.
- W2913505554 hasConcept C154945302 @default.
- W2913505554 hasConcept C162324750 @default.
- W2913505554 hasConcept C176217482 @default.
- W2913505554 hasConcept C1921717 @default.
- W2913505554 hasConcept C205203396 @default.
- W2913505554 hasConcept C21547014 @default.
- W2913505554 hasConcept C31972630 @default.
- W2913505554 hasConcept C33923547 @default.
- W2913505554 hasConcept C41008148 @default.
- W2913505554 hasConcept C774472 @default.
- W2913505554 hasConcept C77553402 @default.
- W2913505554 hasConceptScore W2913505554C103278499 @default.
- W2913505554 hasConceptScore W2913505554C115961682 @default.
- W2913505554 hasConceptScore W2913505554C119857082 @default.
- W2913505554 hasConceptScore W2913505554C134306372 @default.
- W2913505554 hasConceptScore W2913505554C154945302 @default.
- W2913505554 hasConceptScore W2913505554C162324750 @default.
- W2913505554 hasConceptScore W2913505554C176217482 @default.
- W2913505554 hasConceptScore W2913505554C1921717 @default.
- W2913505554 hasConceptScore W2913505554C205203396 @default.
- W2913505554 hasConceptScore W2913505554C21547014 @default.
- W2913505554 hasConceptScore W2913505554C31972630 @default.
- W2913505554 hasConceptScore W2913505554C33923547 @default.
- W2913505554 hasConceptScore W2913505554C41008148 @default.
- W2913505554 hasConceptScore W2913505554C774472 @default.
- W2913505554 hasConceptScore W2913505554C77553402 @default.
- W2913505554 hasLocation W29135055541 @default.
- W2913505554 hasOpenAccess W2913505554 @default.
- W2913505554 hasPrimaryLocation W29135055541 @default.
- W2913505554 hasRelatedWork W1185609327 @default.
- W2913505554 hasRelatedWork W1509176144 @default.
- W2913505554 hasRelatedWork W1965701254 @default.
- W2913505554 hasRelatedWork W2097759558 @default.
- W2913505554 hasRelatedWork W2151880888 @default.
- W2913505554 hasRelatedWork W2173638297 @default.
- W2913505554 hasRelatedWork W2524509220 @default.
- W2913505554 hasRelatedWork W2900438480 @default.
- W2913505554 hasRelatedWork W2934987099 @default.
- W2913505554 hasRelatedWork W3140899861 @default.
- W2913505554 isParatext "false" @default.
- W2913505554 isRetracted "false" @default.
- W2913505554 magId "2913505554" @default.
- W2913505554 workType "article" @default.