Matches in SemOpenAlex for { <https://semopenalex.org/work/W4286903243> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W4286903243 abstract "We present a direct simultaneous speech-to-speech translation (Simul-S2ST) model, Furthermore, the generation of translation is independent from intermediate text representations. Our approach leverages recent progress on direct speech-to-speech translation with discrete units, in which a sequence of discrete representations, instead of continuous spectrogram features, learned in an unsupervised manner, are predicted from the model and passed directly to a vocoder for speech synthesis on-the-fly. We also introduce the variational monotonic multihead attention (V-MMA), to handle the challenge of inefficient policy learning in speech simultaneous translation. The simultaneous policy then operates on source speech features and target discrete units. We carry out empirical studies to compare cascaded and direct approach on the Fisher Spanish-English and MuST-C English-Spanish datasets. Direct simultaneous model is shown to outperform the cascaded model by achieving a better tradeoff between translation quality and latency." @default.
- W4286903243 created "2022-07-25" @default.
- W4286903243 creator A5046819080 @default.
- W4286903243 creator A5051950818 @default.
- W4286903243 creator A5052049405 @default.
- W4286903243 creator A5054878003 @default.
- W4286903243 creator A5058915697 @default.
- W4286903243 creator A5072019513 @default.
- W4286903243 creator A5076306018 @default.
- W4286903243 creator A5077253389 @default.
- W4286903243 creator A5084509956 @default.
- W4286903243 date "2021-10-15" @default.
- W4286903243 modified "2023-10-18" @default.
- W4286903243 title "Direct Simultaneous Speech-to-Speech Translation with Variational Monotonic Multihead Attention" @default.
- W4286903243 doi "https://doi.org/10.48550/arxiv.2110.08250" @default.
- W4286903243 hasPublicationYear "2021" @default.
- W4286903243 type Work @default.
- W4286903243 citedByCount "0" @default.
- W4286903243 crossrefType "posted-content" @default.
- W4286903243 hasAuthorship W4286903243A5046819080 @default.
- W4286903243 hasAuthorship W4286903243A5051950818 @default.
- W4286903243 hasAuthorship W4286903243A5052049405 @default.
- W4286903243 hasAuthorship W4286903243A5054878003 @default.
- W4286903243 hasAuthorship W4286903243A5058915697 @default.
- W4286903243 hasAuthorship W4286903243A5072019513 @default.
- W4286903243 hasAuthorship W4286903243A5076306018 @default.
- W4286903243 hasAuthorship W4286903243A5077253389 @default.
- W4286903243 hasAuthorship W4286903243A5084509956 @default.
- W4286903243 hasBestOaLocation W42869032431 @default.
- W4286903243 hasConcept C104317684 @default.
- W4286903243 hasConcept C105580179 @default.
- W4286903243 hasConcept C134306372 @default.
- W4286903243 hasConcept C149364088 @default.
- W4286903243 hasConcept C14999030 @default.
- W4286903243 hasConcept C154945302 @default.
- W4286903243 hasConcept C185592680 @default.
- W4286903243 hasConcept C203005215 @default.
- W4286903243 hasConcept C204321447 @default.
- W4286903243 hasConcept C2780366754 @default.
- W4286903243 hasConcept C28490314 @default.
- W4286903243 hasConcept C33923547 @default.
- W4286903243 hasConcept C41008148 @default.
- W4286903243 hasConcept C45273575 @default.
- W4286903243 hasConcept C55493867 @default.
- W4286903243 hasConcept C61328038 @default.
- W4286903243 hasConcept C72169020 @default.
- W4286903243 hasConcept C76155785 @default.
- W4286903243 hasConcept C82876162 @default.
- W4286903243 hasConceptScore W4286903243C104317684 @default.
- W4286903243 hasConceptScore W4286903243C105580179 @default.
- W4286903243 hasConceptScore W4286903243C134306372 @default.
- W4286903243 hasConceptScore W4286903243C149364088 @default.
- W4286903243 hasConceptScore W4286903243C14999030 @default.
- W4286903243 hasConceptScore W4286903243C154945302 @default.
- W4286903243 hasConceptScore W4286903243C185592680 @default.
- W4286903243 hasConceptScore W4286903243C203005215 @default.
- W4286903243 hasConceptScore W4286903243C204321447 @default.
- W4286903243 hasConceptScore W4286903243C2780366754 @default.
- W4286903243 hasConceptScore W4286903243C28490314 @default.
- W4286903243 hasConceptScore W4286903243C33923547 @default.
- W4286903243 hasConceptScore W4286903243C41008148 @default.
- W4286903243 hasConceptScore W4286903243C45273575 @default.
- W4286903243 hasConceptScore W4286903243C55493867 @default.
- W4286903243 hasConceptScore W4286903243C61328038 @default.
- W4286903243 hasConceptScore W4286903243C72169020 @default.
- W4286903243 hasConceptScore W4286903243C76155785 @default.
- W4286903243 hasConceptScore W4286903243C82876162 @default.
- W4286903243 hasLocation W42869032431 @default.
- W4286903243 hasOpenAccess W4286903243 @default.
- W4286903243 hasPrimaryLocation W42869032431 @default.
- W4286903243 hasRelatedWork W2102118225 @default.
- W4286903243 hasRelatedWork W2154135679 @default.
- W4286903243 hasRelatedWork W2251328660 @default.
- W4286903243 hasRelatedWork W2587623425 @default.
- W4286903243 hasRelatedWork W3045878282 @default.
- W4286903243 hasRelatedWork W3100608856 @default.
- W4286903243 hasRelatedWork W3207738474 @default.
- W4286903243 hasRelatedWork W4206534379 @default.
- W4286903243 hasRelatedWork W4312706341 @default.
- W4286903243 hasRelatedWork W2178499706 @default.
- W4286903243 isParatext "false" @default.
- W4286903243 isRetracted "false" @default.
- W4286903243 workType "article" @default.