A sub-word based speaker independent speech recognizer using a two-pass segmentation scheme.

Magne Hallstein Johnsen

Browse the full ICASSP paper archive.