Matches in SemOpenAlex for { <https://semopenalex.org/work/W3035237998> ?p ?o ?g. }
- W3035237998 abstract "Generating multi-sentence descriptions for videos is one of the most challenging captioning tasks due to its high requirements for not only visual relevance but also discourse-based coherence across the sentences in the paragraph. Towards this goal, we propose a new approach called Memory-Augmented Recurrent Transformer (MART), which uses a memory module to augment the transformer architecture. The memory module generates a highly summarized memory state from the video segments and the sentence history so as to help better prediction of the next sentence (w.r.t. coreference and repetition aspects), thus encouraging coherent paragraph generation. Extensive experiments, human evaluations, and qualitative analyses on two popular datasets ActivityNet Captions and YouCookII show that MART generates more coherent and less repetitive paragraph captions than baseline methods, while maintaining relevance to the input video events. All code is available open-source at: https://github.com/jayleicn/recurrent-transformer" @default.
- W3035237998 created "2020-06-19" @default.
- W3035237998 creator A5001987532 @default.
- W3035237998 creator A5007285444 @default.
- W3035237998 creator A5008309880 @default.
- W3035237998 creator A5034476404 @default.
- W3035237998 creator A5043356063 @default.
- W3035237998 creator A5055723755 @default.
- W3035237998 date "2020-01-01" @default.
- W3035237998 modified "2023-10-13" @default.
- W3035237998 title "MART: Memory-Augmented Recurrent Transformer for Coherent Video Paragraph Captioning" @default.
- W3035237998 cites W1596841185 @default.
- W3035237998 cites W1815076433 @default.
- W3035237998 cites W1836465849 @default.
- W3035237998 cites W1924770834 @default.
- W3035237998 cites W1927052826 @default.
- W3035237998 cites W1956340063 @default.
- W3035237998 cites W2064675550 @default.
- W3035237998 cites W2101105183 @default.
- W3035237998 cites W2133459682 @default.
- W3035237998 cites W2164290393 @default.
- W3035237998 cites W2194775991 @default.
- W3035237998 cites W2486996822 @default.
- W3035237998 cites W2784025607 @default.
- W3035237998 cites W2883910824 @default.
- W3035237998 cites W2891939431 @default.
- W3035237998 cites W2897439619 @default.
- W3035237998 cites W2899771611 @default.
- W3035237998 cites W2950527759 @default.
- W3035237998 cites W2962937869 @default.
- W3035237998 cites W2963341956 @default.
- W3035237998 cites W2963351113 @default.
- W3035237998 cites W2963403868 @default.
- W3035237998 cites W2963579811 @default.
- W3035237998 cites W2963811641 @default.
- W3035237998 cites W2963916161 @default.
- W3035237998 cites W2964110616 @default.
- W3035237998 cites W2964121744 @default.
- W3035237998 cites W2968101724 @default.
- W3035237998 cites W2970231061 @default.
- W3035237998 cites W2970597249 @default.
- W3035237998 cites W2975501350 @default.
- W3035237998 cites W2981851019 @default.
- W3035237998 cites W2989322838 @default.
- W3035237998 cites W1525783482 @default.
- W3035237998 doi "https://doi.org/10.18653/v1/2020.acl-main.233" @default.
- W3035237998 hasPublicationYear "2020" @default.
- W3035237998 type Work @default.
- W3035237998 sameAs 3035237998 @default.
- W3035237998 citedByCount "91" @default.
- W3035237998 countsByYear W30352379982020 @default.
- W3035237998 countsByYear W30352379982021 @default.
- W3035237998 countsByYear W30352379982022 @default.
- W3035237998 countsByYear W30352379982023 @default.
- W3035237998 crossrefType "proceedings-article" @default.
- W3035237998 hasAuthorship W3035237998A5001987532 @default.
- W3035237998 hasAuthorship W3035237998A5007285444 @default.
- W3035237998 hasAuthorship W3035237998A5008309880 @default.
- W3035237998 hasAuthorship W3035237998A5034476404 @default.
- W3035237998 hasAuthorship W3035237998A5043356063 @default.
- W3035237998 hasAuthorship W3035237998A5055723755 @default.
- W3035237998 hasBestOaLocation W30352379981 @default.
- W3035237998 hasConcept C111919701 @default.
- W3035237998 hasConcept C115961682 @default.
- W3035237998 hasConcept C118505674 @default.
- W3035237998 hasConcept C119599485 @default.
- W3035237998 hasConcept C127413603 @default.
- W3035237998 hasConcept C136764020 @default.
- W3035237998 hasConcept C154945302 @default.
- W3035237998 hasConcept C157657479 @default.
- W3035237998 hasConcept C165801399 @default.
- W3035237998 hasConcept C204321447 @default.
- W3035237998 hasConcept C2777206241 @default.
- W3035237998 hasConcept C2777530160 @default.
- W3035237998 hasConcept C28490314 @default.
- W3035237998 hasConcept C41008148 @default.
- W3035237998 hasConcept C66322947 @default.
- W3035237998 hasConceptScore W3035237998C111919701 @default.
- W3035237998 hasConceptScore W3035237998C115961682 @default.
- W3035237998 hasConceptScore W3035237998C118505674 @default.
- W3035237998 hasConceptScore W3035237998C119599485 @default.
- W3035237998 hasConceptScore W3035237998C127413603 @default.
- W3035237998 hasConceptScore W3035237998C136764020 @default.
- W3035237998 hasConceptScore W3035237998C154945302 @default.
- W3035237998 hasConceptScore W3035237998C157657479 @default.
- W3035237998 hasConceptScore W3035237998C165801399 @default.
- W3035237998 hasConceptScore W3035237998C204321447 @default.
- W3035237998 hasConceptScore W3035237998C2777206241 @default.
- W3035237998 hasConceptScore W3035237998C2777530160 @default.
- W3035237998 hasConceptScore W3035237998C28490314 @default.
- W3035237998 hasConceptScore W3035237998C41008148 @default.
- W3035237998 hasConceptScore W3035237998C66322947 @default.
- W3035237998 hasLocation W30352379981 @default.
- W3035237998 hasLocation W30352379982 @default.
- W3035237998 hasOpenAccess W3035237998 @default.
- W3035237998 hasPrimaryLocation W30352379981 @default.
- W3035237998 hasRelatedWork W10582454 @default.
- W3035237998 hasRelatedWork W12873135 @default.
- W3035237998 hasRelatedWork W1453065 @default.
- W3035237998 hasRelatedWork W14719935 @default.