Sequence-to-sequence models for punctuated transcription combining lexical and acoustic features.

Ondrej Klejch, Peter Bell, Steve Renals

Browse the full ICASSP paper archive.