A tail-sharing WFST composition algorithm for large vocabulary speech recognition.

Diamantino Caseiro, Isabel Trancoso

Browse the full ICASSP paper archive.