Listen and Fill in the Missing Letters: Non-Autoregressive Transformer for Speech Recognition.
Nanxin ChenShinji WatanabeJesús VillalbaNajim DehakPublished in: CoRR (2019)
Keyphrases
- speech recognition
- autoregressive
- non stationary
- hidden markov models
- gaussian markov random field
- automatic speech recognition
- speech synthesis
- random fields
- speech processing
- spectral analysis
- language model
- speech recognizer
- pattern recognition
- speech recognition technology
- speech signal
- speech recognition systems
- linear prediction
- noisy environments
- sar images
- multiresolution
- neural network
- speaker identification
- semi supervised