G2G: TTS-Driven Pronunciation Learning for Graphemic Hybrid ASR.

Duc Le, Thilo Khler, Christian Fuegen, Michael L. Seltzer

Browse the full ICASSP paper archive.