Cycle-consistency training for end-to-end speech recognition.
Takaaki HoriRamón Fernandez AstudilloTomoki HayashiYu ZhangShinji WatanabeJonathan Le RouxPublished in: CoRR (2018)
Keyphrases
- end to end
- speech recognition
- wall street journal corpus
- isolated word
- hidden markov models
- language model
- acoustic models
- speech processing
- automatic speech recognition
- speech synthesis
- speech recognition technology
- speech signal
- noisy environments
- pattern recognition
- congestion control
- speaker identification
- speech recognizer
- speaker independent
- speech retrieval
- speech recognition systems
- speaker dependent
- discriminative training
- speaker adaptation
- neural network
- speech recognizers
- feature extraction
- feature selection