Training Autoregressive Speech Recognition Models with Limited in-domain Supervision.
Chak-Fai LiFrancis KeithWilliam HartmannMatthew SnoverPublished in: CoRR (2022)
Keyphrases
- autoregressive
- speech recognition
- acoustic models
- non stationary
- random fields
- moving average
- hidden markov models
- random field models
- pattern recognition
- speech signal
- automatic speech recognition
- wall street journal corpus
- gaussian markov random field
- speech synthesis
- language model
- speech recognizer
- speaker independent
- noisy environments
- spectral analysis
- speech recognition systems
- sar images
- autoregressive moving average
- similarity measure