A connected speech recognition system using a diphone-based language model.

Anna Maria Colla, Carlo Scagliola, Donatella Sciarra

Browse the full ICASSP paper archive.