Context-dependent acoustic modeling using graphemes for large vocabulary speech recognition.

Stephan Kanthak, Hermann Ney

Browse the full ICASSP paper archive.