ICASSP
2009
IEEE
14 years 6 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...