duchateau05@interspeech_2005@ISCA

Total: 1

#1 Statistical language models for large vocabulary spontaneous speech recognition in dutch [PDF] [Copy] [Kimi] [REL]

Authors: Jacques Duchateau, Dong Hoon Van Uytsel, Hugo Van Hamme, Patrick Wambacq

In state-of-the-art large vocabulary automatic recognition systems, a large statistical language model is used, typically an N-gram. However in order to estimate this model, a large database of sentences or texts in the same style as the recognition task is needed. For spontaneous speech one doesn't dispose of such database since it should consist of accurate thus expensive orthographic transcriptions of spoken audio.