Non-Autoregressive TTS with Explicit Duration Modelling for Low-Resource Highly Expressive Speech.
Raahil ShahKamil PokoraAbdelhamid EzzergViacheslav KlimkovGoeric HuybrechtsBartosz PutryczDaniel KorzekwaThomas MerrittPublished in: SSW (2021)
Keyphrases
- autoregressive
- text to speech
- highly expressive
- non stationary
- moving average
- gaussian markov random field
- random fields
- first order logic
- description logics
- speech signal
- speech recognition
- predicate logic
- machine learning
- random field models
- knowledge representation
- reasoning systems
- defeasible logic
- temporal logic