Improved acoustic modeling for speech recognition using 2D Markov random fields.

Helmut Lucke

Browse the full ICASSP paper archive.