On Using 2D Sequence-to-sequence Models for Speech Recognition.

Parnia Bahar, Albert Zeyer, Ralf Schlter, Hermann Ney

Browse the full ICASSP paper archive.