Leveraging large amounts of loosely transcribed corporate videos for acoustic model training.

Matthias Paulik, Panchi Panchapagesan

Browse the full ASRU paper archive.