Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136875705> ?p ?o ?g. }
Showing items 1 to 78 of
78
with 100 items per page.
- W3136875705 endingPage "80" @default.
- W3136875705 startingPage "72" @default.
- W3136875705 abstract "Dense video captioning aims to generate dense descriptions for all possible events in an untrimmed video. The task is challenging that it requires accurately localizing events in the video and simultaneously describe each event with a sentence. Current approaches usually decompose this task into two independent stages—the proposal localization stage and the caption generation stage, resulting in a suboptimal solution. Masked Transformer (MT) model [30] has been proposed to integrate the two stages and optimize them in an end-to-end philosophy. Despite the superior performance that the MT has achieved, its runtime efficiency is unsatisfactory which severely limits its applicability in real-world scenarios. In this paper, we devise an improved Accelerated Masked Transformer (AMT) model that enjoys the dual-benefit of effectiveness and efficiency. Taking MT as our reference model, we respectively introduce accelerating strategies to the two stages: 1) in the proposal localization stage, we introduce a lightweight anchor-free proposal in company with a local attention mechanism; and 2) in the caption generation stage, we introduce the single-shot feature masking strategy along with an average attention mechanism. Extensive experiments on two benchmark datasets ActivityNet-Caption and YouCookII demonstrate that AMT achieves competitive performance on both datasets with significant speed improvement. On the ActivityNet-Caption dataset, AMT reduces up to 2× running time with comparable performance when compared to the reference MT model." @default.
- W3136875705 created "2021-03-29" @default.
- W3136875705 creator A5055626738 @default.
- W3136875705 creator A5061025828 @default.
- W3136875705 date "2021-07-01" @default.
- W3136875705 modified "2023-10-14" @default.
- W3136875705 title "Accelerated masked transformer for dense video captioning" @default.
- W3136875705 cites W1996219872 @default.
- W3136875705 cites W2808181286 @default.
- W3136875705 cites W2952132648 @default.
- W3136875705 cites W2963150162 @default.
- W3136875705 cites W2964213727 @default.
- W3136875705 cites W2981165461 @default.
- W3136875705 doi "https://doi.org/10.1016/j.neucom.2021.03.026" @default.
- W3136875705 hasPublicationYear "2021" @default.
- W3136875705 type Work @default.
- W3136875705 sameAs 3136875705 @default.
- W3136875705 citedByCount "8" @default.
- W3136875705 countsByYear W31368757052022 @default.
- W3136875705 countsByYear W31368757052023 @default.
- W3136875705 crossrefType "journal-article" @default.
- W3136875705 hasAuthorship W3136875705A5055626738 @default.
- W3136875705 hasAuthorship W3136875705A5061025828 @default.
- W3136875705 hasConcept C115961682 @default.
- W3136875705 hasConcept C121332964 @default.
- W3136875705 hasConcept C13280743 @default.
- W3136875705 hasConcept C154945302 @default.
- W3136875705 hasConcept C157657479 @default.
- W3136875705 hasConcept C162324750 @default.
- W3136875705 hasConcept C165801399 @default.
- W3136875705 hasConcept C185798385 @default.
- W3136875705 hasConcept C187736073 @default.
- W3136875705 hasConcept C205649164 @default.
- W3136875705 hasConcept C2777530160 @default.
- W3136875705 hasConcept C2780451532 @default.
- W3136875705 hasConcept C28490314 @default.
- W3136875705 hasConcept C41008148 @default.
- W3136875705 hasConcept C62520636 @default.
- W3136875705 hasConcept C66322947 @default.
- W3136875705 hasConcept C79403827 @default.
- W3136875705 hasConceptScore W3136875705C115961682 @default.
- W3136875705 hasConceptScore W3136875705C121332964 @default.
- W3136875705 hasConceptScore W3136875705C13280743 @default.
- W3136875705 hasConceptScore W3136875705C154945302 @default.
- W3136875705 hasConceptScore W3136875705C157657479 @default.
- W3136875705 hasConceptScore W3136875705C162324750 @default.
- W3136875705 hasConceptScore W3136875705C165801399 @default.
- W3136875705 hasConceptScore W3136875705C185798385 @default.
- W3136875705 hasConceptScore W3136875705C187736073 @default.
- W3136875705 hasConceptScore W3136875705C205649164 @default.
- W3136875705 hasConceptScore W3136875705C2777530160 @default.
- W3136875705 hasConceptScore W3136875705C2780451532 @default.
- W3136875705 hasConceptScore W3136875705C28490314 @default.
- W3136875705 hasConceptScore W3136875705C41008148 @default.
- W3136875705 hasConceptScore W3136875705C62520636 @default.
- W3136875705 hasConceptScore W3136875705C66322947 @default.
- W3136875705 hasConceptScore W3136875705C79403827 @default.
- W3136875705 hasFunder F4320321001 @default.
- W3136875705 hasLocation W31368757051 @default.
- W3136875705 hasOpenAccess W3136875705 @default.
- W3136875705 hasPrimaryLocation W31368757051 @default.
- W3136875705 hasRelatedWork W2889111319 @default.
- W3136875705 hasRelatedWork W2891955747 @default.
- W3136875705 hasRelatedWork W2949789546 @default.
- W3136875705 hasRelatedWork W3025136821 @default.
- W3136875705 hasRelatedWork W3035237998 @default.
- W3136875705 hasRelatedWork W3038899388 @default.
- W3136875705 hasRelatedWork W3209355071 @default.
- W3136875705 hasRelatedWork W3216250699 @default.
- W3136875705 hasRelatedWork W4223519544 @default.
- W3136875705 hasRelatedWork W4361193378 @default.
- W3136875705 hasVolume "445" @default.
- W3136875705 isParatext "false" @default.
- W3136875705 isRetracted "false" @default.
- W3136875705 magId "3136875705" @default.
- W3136875705 workType "article" @default.