Semi-supervised Training for End-to-end Models via Weak Distillation.

Bo Li, Tara N. Sainath, Ruoming Pang, Zelin Wu

Browse the full ICASSP paper archive.