Non-Autoregressive End-To-End Automatic Speech Recognition Incorporating Downstream Natural Language Processing.
Motoi OmachiYuya FujitaShinji WatanabeTianzi WangPublished in: ICASSP (2022)
Keyphrases
- end to end
- autoregressive
- automatic speech recognition
- natural language processing
- speech recognition
- non stationary
- speech signal
- moving average
- gaussian markov random field
- hidden markov models
- information extraction
- word error rate
- speech retrieval
- machine learning
- conversational speech
- broadcast news
- random fields
- natural language
- congestion control
- machine translation
- expectation maximization
- multiresolution
- information retrieval