Sequence Noise Injected Training for End-to-end Speech Recognition.

George Saon, Zoltn Tske, Kartik Audhkhasi, Brian Kingsbury

Browse the full ICASSP paper archive.