ICASSP
2009
IEEE
13 years 11 months ago
2009
IEEE
For TV and radio shows containing narrowband speech, Speech-to-text (STT) accuracy on the narrowband audio can be improved by using an acoustic model trained on acoustically match...