Incremental Text-to-Speech Synthesis Using Pseudo Lookahead With Large Pretrained Language Model
Sign Up to like & getrecommendations! Published in 2021 at "IEEE Signal Processing Letters"
DOI: 10.1109/lsp.2021.3073869
Abstract: This letter presents an incremental text-to-speech (TTS) method that performs synthesis in small linguistic units while maintaining the naturalness of output speech. Incremental TTS is generally subject to a trade-off between latency and synthetic speech… read more here.
Keywords: language model; pseudo lookahead; method; text speech ... See more keywords