Four-in-One: a Joint Approach to Inverse Text Normalization, Punctuation, Capitalization, and Disfluency for Automatic Speech Recognition.
Sharman TanPiyush BehreNick KibreIssac AlphonsoShuangyu ChangPublished in: SLT (2022)
Keyphrases
- automatic speech recognition
- spontaneous speech
- conversational speech
- speech recognition
- spoken words
- hidden markov models
- speech retrieval
- human machine interaction
- speech signal
- spoken language
- word error rate
- broadcast news
- english text
- spoken document retrieval
- noisy environments
- recognition errors
- multiscale
- compound words
- text mining
- length normalization