Matches in SemOpenAlex for { <https://semopenalex.org/work/W3014704188> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W3014704188 abstract "Scaling up the vocabulary and complexity of current visual understanding systems is necessary in order to bridge the gap between human and machine visual intelligence. However, a crucial impediment to this end lies in the difficulty of generalizing to data distributions that come from real-world scenarios. Typically such distributions follow Zipf's law which states that only a small portion of the collected object classes will have abundant examples (head); while most classes will contain just a few (tail). In this paper, we propose to study a novel task concerning the generalization of visual relationships that are on the distribution's tail, i.e. we investigate how to help AI systems to better recognize rare relationships like , where the subject S, predicate P, and/or the object O come from the tail of the corresponding distributions. To achieve this goal, we first introduce two large-scale visual-relationship detection benchmarks built upon the widely used Visual Genome and GQA datasets. We also propose an intuitive evaluation protocol that gives credit to classifiers who prefer concepts that are semantically close to the ground truth class according to wordNet- or word2vec-induced metrics. Finally, we introduce a visiolinguistic version of a Hubless loss which we show experimentally that it consistently encourages classifiers to be more predictive of the tail classes while still being accurate on head classes. Our code and models are available on this http URL." @default.
- W3014704188 created "2020-04-10" @default.
- W3014704188 creator A5016543371 @default.
- W3014704188 creator A5036945471 @default.
- W3014704188 creator A5038007540 @default.
- W3014704188 creator A5065154024 @default.
- W3014704188 creator A5082470433 @default.
- W3014704188 creator A5085089542 @default.
- W3014704188 date "2020-03-25" @default.
- W3014704188 modified "2023-09-30" @default.
- W3014704188 title "Long-tail Visual Relationship Recognition with a Visiolinguistic Hubless Loss" @default.
- W3014704188 hasPublicationYear "2020" @default.
- W3014704188 type Work @default.
- W3014704188 sameAs 3014704188 @default.
- W3014704188 citedByCount "6" @default.
- W3014704188 countsByYear W30147041882020 @default.
- W3014704188 countsByYear W30147041882021 @default.
- W3014704188 crossrefType "posted-content" @default.
- W3014704188 hasAuthorship W3014704188A5016543371 @default.
- W3014704188 hasAuthorship W3014704188A5036945471 @default.
- W3014704188 hasAuthorship W3014704188A5038007540 @default.
- W3014704188 hasAuthorship W3014704188A5065154024 @default.
- W3014704188 hasAuthorship W3014704188A5082470433 @default.
- W3014704188 hasAuthorship W3014704188A5085089542 @default.
- W3014704188 hasConcept C105795698 @default.
- W3014704188 hasConcept C119857082 @default.
- W3014704188 hasConcept C125932096 @default.
- W3014704188 hasConcept C130318100 @default.
- W3014704188 hasConcept C134306372 @default.
- W3014704188 hasConcept C140146324 @default.
- W3014704188 hasConcept C154945302 @default.
- W3014704188 hasConcept C157659113 @default.
- W3014704188 hasConcept C177148314 @default.
- W3014704188 hasConcept C199360897 @default.
- W3014704188 hasConcept C204321447 @default.
- W3014704188 hasConcept C2776461190 @default.
- W3014704188 hasConcept C2778828372 @default.
- W3014704188 hasConcept C33923547 @default.
- W3014704188 hasConcept C41008148 @default.
- W3014704188 hasConcept C41608201 @default.
- W3014704188 hasConceptScore W3014704188C105795698 @default.
- W3014704188 hasConceptScore W3014704188C119857082 @default.
- W3014704188 hasConceptScore W3014704188C125932096 @default.
- W3014704188 hasConceptScore W3014704188C130318100 @default.
- W3014704188 hasConceptScore W3014704188C134306372 @default.
- W3014704188 hasConceptScore W3014704188C140146324 @default.
- W3014704188 hasConceptScore W3014704188C154945302 @default.
- W3014704188 hasConceptScore W3014704188C157659113 @default.
- W3014704188 hasConceptScore W3014704188C177148314 @default.
- W3014704188 hasConceptScore W3014704188C199360897 @default.
- W3014704188 hasConceptScore W3014704188C204321447 @default.
- W3014704188 hasConceptScore W3014704188C2776461190 @default.
- W3014704188 hasConceptScore W3014704188C2778828372 @default.
- W3014704188 hasConceptScore W3014704188C33923547 @default.
- W3014704188 hasConceptScore W3014704188C41008148 @default.
- W3014704188 hasConceptScore W3014704188C41608201 @default.
- W3014704188 hasLocation W30147041881 @default.
- W3014704188 hasOpenAccess W3014704188 @default.
- W3014704188 hasPrimaryLocation W30147041881 @default.
- W3014704188 hasRelatedWork W1937075317 @default.
- W3014704188 hasRelatedWork W2250804920 @default.
- W3014704188 hasRelatedWork W2251390968 @default.
- W3014704188 hasRelatedWork W2340492831 @default.
- W3014704188 hasRelatedWork W2418478089 @default.
- W3014704188 hasRelatedWork W2617070002 @default.
- W3014704188 hasRelatedWork W2784977129 @default.
- W3014704188 hasRelatedWork W2850386316 @default.
- W3014704188 hasRelatedWork W2896659472 @default.
- W3014704188 hasRelatedWork W2903359200 @default.
- W3014704188 hasRelatedWork W2945003880 @default.
- W3014704188 hasRelatedWork W2989208022 @default.
- W3014704188 hasRelatedWork W3028770966 @default.
- W3014704188 hasRelatedWork W3093264153 @default.
- W3014704188 hasRelatedWork W3105511442 @default.
- W3014704188 hasRelatedWork W3119636502 @default.
- W3014704188 hasRelatedWork W3125115207 @default.
- W3014704188 hasRelatedWork W3132516639 @default.
- W3014704188 hasRelatedWork W3174167978 @default.
- W3014704188 hasRelatedWork W3198734150 @default.
- W3014704188 isParatext "false" @default.
- W3014704188 isRetracted "false" @default.
- W3014704188 magId "3014704188" @default.
- W3014704188 workType "article" @default.