Building competitive direct acoustics-to-word models for English conversational speech recognition.
Kartik AudhkhasiBrian KingsburyBhuvana RamabhadranGeorge SaonMichael PichenyPublished in: CoRR (2017)
Keyphrases
- speech recognition
- speech recognizer
- speech recognition systems
- speech recognition technology
- hidden markov models
- automatic speech recognition
- pattern recognition
- keyword spotting
- speech processing
- acoustic models
- probabilistic model
- wall street journal corpus
- speech signal
- language model
- conversational speech
- speech recognizers
- spoken language
- speech retrieval
- noisy environments
- speech synthesis
- natural language
- english text
- speaker identification
- broadcast news
- machine translation
- n gram
- mobile devices
- word error rate
- machine translation system
- speaker independent
- english language
- translation model
- speaker adaptation
- information extraction