Investigating the Utility of Surprisal from Large Language Models for Speech Synthesis Prosody.
Sofoklis KakourosJuraj SimkoMartti VainioAntti SuniPublished in: CoRR (2023)
Keyphrases
- speech synthesis
- language model
- speech recognition
- language modeling
- vocal tract
- probabilistic model
- prosodic features
- text to speech
- n gram
- document retrieval
- information retrieval
- speech signal
- automatic speech recognition
- query expansion
- context sensitive
- language modelling
- query terms
- statistical language models
- document ranking
- smoothing methods
- test collection
- retrieval model
- ad hoc information retrieval
- okapi bm
- translation model
- vector space model
- error rate
- passage retrieval
- pseudo relevance feedback
- neural network