Towards better performance with heterogeneous training data in acoustic modeling using deep neural networks.

Yan Huang, Malcolm Slaney, Michael L. Seltzer, Yifan Gong

Browse the full Interspeech paper archive.