Pseudo-segment based speech recognition using neural recurrent whole-word recognizers.

Philippe Le Cerf, Kris Demuynck, Jacques Duchateau, Dirk Van Compernolle

Browse the full ICASSP paper archive.