Skip to main content
Publication

Building context-dependent DNN acoustic models using Kullback-Leibler divergence-based state tying