Acoustic-to-Word Attention-Based Model Complemented with Character-Level CTC-Based Model.

Sei Ueno, Hirofumi Inaguma, Masato Mimura, Tatsuya Kawahara

Browse the full ICASSP paper archive.