Matches in SemOpenAlex for { <https://semopenalex.org/work/W4288805336> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W4288805336 abstract "End-to-end speech synthesis models directly convert the input characters into an audio representation (e.g., spectrograms). Despite their impressive performance, such models have difficulty disambiguating the pronunciations of identically spelled words. To mitigate this issue, a separate Grapheme-to-Phoneme (G2P) model can be employed to convert the characters into phonemes before synthesizing the audio. This paper proposes SoundChoice, a novel G2P architecture that processes entire sentences rather than operating at the word level. The proposed architecture takes advantage of a weighted homograph loss (that improves disambiguation), exploits curriculum learning (that gradually switches from word-level to sentence-level G2P), and integrates word embeddings from BERT (for further performance improvement). Moreover, the model inherits the best practices in speech recognition, including multi-task learning with Connectionist Temporal Classification (CTC) and beam search with an embedded language model. As a result, SoundChoice achieves a Phoneme Error Rate (PER) of 2.65% on whole-sentence transcription using data from LibriSpeech and Wikipedia. Index Terms grapheme-to-phoneme, speech synthesis, text-tospeech, phonetics, pronunciation, disambiguation." @default.
- W4288805336 created "2022-07-30" @default.
- W4288805336 creator A5040811098 @default.
- W4288805336 creator A5063047435 @default.
- W4288805336 date "2022-07-26" @default.
- W4288805336 modified "2023-09-27" @default.
- W4288805336 title "SoundChoice: Grapheme-to-Phoneme Models with Semantic Disambiguation" @default.
- W4288805336 doi "https://doi.org/10.48550/arxiv.2207.13703" @default.
- W4288805336 hasPublicationYear "2022" @default.
- W4288805336 type Work @default.
- W4288805336 citedByCount "0" @default.
- W4288805336 crossrefType "posted-content" @default.
- W4288805336 hasAuthorship W4288805336A5040811098 @default.
- W4288805336 hasAuthorship W4288805336A5063047435 @default.
- W4288805336 hasBestOaLocation W42888053361 @default.
- W4288805336 hasConcept C121332964 @default.
- W4288805336 hasConcept C137293760 @default.
- W4288805336 hasConcept C138885662 @default.
- W4288805336 hasConcept C14999030 @default.
- W4288805336 hasConcept C154945302 @default.
- W4288805336 hasConcept C179926584 @default.
- W4288805336 hasConcept C204321447 @default.
- W4288805336 hasConcept C2776779415 @default.
- W4288805336 hasConcept C2777530160 @default.
- W4288805336 hasConcept C2780844864 @default.
- W4288805336 hasConcept C28490314 @default.
- W4288805336 hasConcept C30080830 @default.
- W4288805336 hasConcept C40969351 @default.
- W4288805336 hasConcept C41008148 @default.
- W4288805336 hasConcept C41895202 @default.
- W4288805336 hasConcept C50644808 @default.
- W4288805336 hasConcept C62520636 @default.
- W4288805336 hasConcept C8521452 @default.
- W4288805336 hasConcept C90805587 @default.
- W4288805336 hasConceptScore W4288805336C121332964 @default.
- W4288805336 hasConceptScore W4288805336C137293760 @default.
- W4288805336 hasConceptScore W4288805336C138885662 @default.
- W4288805336 hasConceptScore W4288805336C14999030 @default.
- W4288805336 hasConceptScore W4288805336C154945302 @default.
- W4288805336 hasConceptScore W4288805336C179926584 @default.
- W4288805336 hasConceptScore W4288805336C204321447 @default.
- W4288805336 hasConceptScore W4288805336C2776779415 @default.
- W4288805336 hasConceptScore W4288805336C2777530160 @default.
- W4288805336 hasConceptScore W4288805336C2780844864 @default.
- W4288805336 hasConceptScore W4288805336C28490314 @default.
- W4288805336 hasConceptScore W4288805336C30080830 @default.
- W4288805336 hasConceptScore W4288805336C40969351 @default.
- W4288805336 hasConceptScore W4288805336C41008148 @default.
- W4288805336 hasConceptScore W4288805336C41895202 @default.
- W4288805336 hasConceptScore W4288805336C50644808 @default.
- W4288805336 hasConceptScore W4288805336C62520636 @default.
- W4288805336 hasConceptScore W4288805336C8521452 @default.
- W4288805336 hasConceptScore W4288805336C90805587 @default.
- W4288805336 hasLocation W42888053361 @default.
- W4288805336 hasOpenAccess W4288805336 @default.
- W4288805336 hasPrimaryLocation W42888053361 @default.
- W4288805336 hasRelatedWork W10658944 @default.
- W4288805336 hasRelatedWork W11920722 @default.
- W4288805336 hasRelatedWork W14754660 @default.
- W4288805336 hasRelatedWork W1745277 @default.
- W4288805336 hasRelatedWork W2308727 @default.
- W4288805336 hasRelatedWork W5457946 @default.
- W4288805336 hasRelatedWork W7606746 @default.
- W4288805336 hasRelatedWork W8053366 @default.
- W4288805336 hasRelatedWork W849475 @default.
- W4288805336 hasRelatedWork W8738421 @default.
- W4288805336 isParatext "false" @default.
- W4288805336 isRetracted "false" @default.
- W4288805336 workType "article" @default.