Login / Signup

Wav2Seq: Pre-training Speech-to-Text Encoder-Decoder Models Using Pseudo Languages.

Felix WuKwangyoun KimShinji WatanabeKyu J. HanRyan McDonaldKilian Q. WeinbergerYoav Artzi
Published in: CoRR (2022)
Keyphrases
  • low complexity
  • motion estimation
  • training set
  • noisy channel
  • probabilistic model
  • expressive power
  • language independent
  • rate allocation