Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897927512> ?p ?o ?g. }
Showing items 1 to 91 of
91
with 100 items per page.
- W2897927512 endingPage "19" @default.
- W2897927512 startingPage "1" @default.
- W2897927512 abstract "Recently, a series of attempts have incorporated spatial attention mechanisms into the task of image captioning, which achieves a remarkable improvement in the quality of generative captions. However, the traditional spatial attention mechanism adopts latent and delayed semantic representations to decide which area should be paid more attention to, resulting in inaccurate semantic guidance and the introduction of redundant information. In order to optimize the spatial attention mechanism, we propose the Semantic Guidance Attention (SGA) mechanism in this article. Specifically, SGA utilizes semantic word representations to provide an intuitive semantic guidance that focuses accurately on semantic-related regions. Moreover, we reduce the difficulty of generating fluent sentences by updating the attention information in time. At the same time, the beam search algorithm is widely used to predict words during sequence generation. This algorithm generates a sentence according to the probabilities of words, so it is easy to push out a generic sentence and discard some distinctive captions. In order to overcome this limitation, we design the Consensus Selection (CS) strategy to choose the most descriptive and informative caption, which is selected by the semantic similarity of captions instead of the probabilities of words. The consensus caption is determined by selecting the one with the highest cumulative semantic similarity with respect to the reference captions. Our proposed model (SGA-CS) is validated on Flickr30k and MSCOCO, which shows that SGA-CS outperforms state-of-the-art approaches. To our best knowledge, SGA-CS is the first attempt to jointly produce semantic attention guidance and select descriptive captions for image captioning tasks, achieving one of the best performance ratings among any cross-entropy training methods." @default.
- W2897927512 created "2018-10-26" @default.
- W2897927512 creator A5008951080 @default.
- W2897927512 creator A5029203943 @default.
- W2897927512 creator A5045089479 @default.
- W2897927512 date "2018-10-10" @default.
- W2897927512 modified "2023-09-26" @default.
- W2897927512 title "Image Captioning via Semantic Guidance Attention and Consensus Selection Strategy" @default.
- W2897927512 cites W1773149199 @default.
- W2897927512 cites W1895577753 @default.
- W2897927512 cites W1905882502 @default.
- W2897927512 cites W1956340063 @default.
- W2897927512 cites W2064675550 @default.
- W2897927512 cites W2101105183 @default.
- W2897927512 cites W2194775991 @default.
- W2897927512 cites W2220981600 @default.
- W2897927512 cites W2302086703 @default.
- W2897927512 cites W2564898401 @default.
- W2897927512 cites W2739107216 @default.
- W2897927512 cites W2766046458 @default.
- W2897927512 cites W2963300078 @default.
- W2897927512 doi "https://doi.org/10.1145/3271485" @default.
- W2897927512 hasPublicationYear "2018" @default.
- W2897927512 type Work @default.
- W2897927512 sameAs 2897927512 @default.
- W2897927512 citedByCount "6" @default.
- W2897927512 countsByYear W28979275122020 @default.
- W2897927512 countsByYear W28979275122021 @default.
- W2897927512 countsByYear W28979275122023 @default.
- W2897927512 crossrefType "journal-article" @default.
- W2897927512 hasAuthorship W2897927512A5008951080 @default.
- W2897927512 hasAuthorship W2897927512A5029203943 @default.
- W2897927512 hasAuthorship W2897927512A5045089479 @default.
- W2897927512 hasConcept C103278499 @default.
- W2897927512 hasConcept C115961682 @default.
- W2897927512 hasConcept C130318100 @default.
- W2897927512 hasConcept C138885662 @default.
- W2897927512 hasConcept C154945302 @default.
- W2897927512 hasConcept C157657479 @default.
- W2897927512 hasConcept C162324750 @default.
- W2897927512 hasConcept C187736073 @default.
- W2897927512 hasConcept C204321447 @default.
- W2897927512 hasConcept C23123220 @default.
- W2897927512 hasConcept C2777530160 @default.
- W2897927512 hasConcept C2780451532 @default.
- W2897927512 hasConcept C41008148 @default.
- W2897927512 hasConcept C41895202 @default.
- W2897927512 hasConcept C67277372 @default.
- W2897927512 hasConcept C81917197 @default.
- W2897927512 hasConcept C90805587 @default.
- W2897927512 hasConceptScore W2897927512C103278499 @default.
- W2897927512 hasConceptScore W2897927512C115961682 @default.
- W2897927512 hasConceptScore W2897927512C130318100 @default.
- W2897927512 hasConceptScore W2897927512C138885662 @default.
- W2897927512 hasConceptScore W2897927512C154945302 @default.
- W2897927512 hasConceptScore W2897927512C157657479 @default.
- W2897927512 hasConceptScore W2897927512C162324750 @default.
- W2897927512 hasConceptScore W2897927512C187736073 @default.
- W2897927512 hasConceptScore W2897927512C204321447 @default.
- W2897927512 hasConceptScore W2897927512C23123220 @default.
- W2897927512 hasConceptScore W2897927512C2777530160 @default.
- W2897927512 hasConceptScore W2897927512C2780451532 @default.
- W2897927512 hasConceptScore W2897927512C41008148 @default.
- W2897927512 hasConceptScore W2897927512C41895202 @default.
- W2897927512 hasConceptScore W2897927512C67277372 @default.
- W2897927512 hasConceptScore W2897927512C81917197 @default.
- W2897927512 hasConceptScore W2897927512C90805587 @default.
- W2897927512 hasFunder F4320321001 @default.
- W2897927512 hasFunder F4320321921 @default.
- W2897927512 hasFunder F4320335787 @default.
- W2897927512 hasIssue "4" @default.
- W2897927512 hasLocation W28979275121 @default.
- W2897927512 hasOpenAccess W2897927512 @default.
- W2897927512 hasPrimaryLocation W28979275121 @default.
- W2897927512 hasRelatedWork W2002382481 @default.
- W2897927512 hasRelatedWork W2096589809 @default.
- W2897927512 hasRelatedWork W2116838603 @default.
- W2897927512 hasRelatedWork W2289318896 @default.
- W2897927512 hasRelatedWork W2349125667 @default.
- W2897927512 hasRelatedWork W2766760871 @default.
- W2897927512 hasRelatedWork W2974225181 @default.
- W2897927512 hasRelatedWork W3078371441 @default.
- W2897927512 hasRelatedWork W4287890973 @default.
- W2897927512 hasRelatedWork W4288108740 @default.
- W2897927512 hasVolume "14" @default.
- W2897927512 isParatext "false" @default.
- W2897927512 isRetracted "false" @default.
- W2897927512 magId "2897927512" @default.
- W2897927512 workType "article" @default.