FastSpeech2

GPTKB entity

Statements (33)
Predicate Object
gptkbp:instanceOf speech synthesis model
gptkbp:application text-to-speech
gptkbp:architecture non-autoregressive
gptkbp:author gptkb:Xu_Tan
gptkb:Tie-Yan_Liu
Sheng Zhao
Tao Qin
Yangjun Ruan
Yi Ren
Zhou Zhao
gptkbp:basedOn FastSpeech
gptkbp:citation high (over 1000)
gptkbp:developedBy gptkb:Microsoft_Research_Asia
https://www.w3.org/2000/01/rdf-schema#label FastSpeech2
gptkbp:improves FastSpeech
gptkbp:input phoneme sequence
gptkbp:introducedIn 2020
gptkbp:language English
gptkbp:notableFeature fast inference speed
explicit prosody modeling
high-quality speech synthesis
parallel generation
gptkbp:notablePublication FastSpeech 2: Fast and High-Quality End-to-End Text to Speech
gptkbp:openSource yes
gptkbp:output mel-spectrogram
gptkbp:predicts gptkb:energy
pitch
duration
gptkbp:publicationDate gptkb:AAAI_2021
gptkbp:url https://arxiv.org/abs/2006.04558
gptkbp:usedFor neural TTS
gptkbp:bfsParent gptkb:Hugging_Face_models
gptkbp:bfsLayer 7