How Neural Network Depth Compensates for HMM Conditional Independence Assumptions in DNN-HMM Acoustic Models

Suman V. Ravuri, Steven Wegmann. How Neural Network Depth Compensates for HMM Conditional Independence Assumptions in DNN-HMM Acoustic Models. In Nelson Morgan, editor, Interspeech 2016, 17th Annual Conference of the International Speech Communication Association, San Francisco, CA, USA, September 8-12, 2016. pages 2736-2740, ISCA, 2016. [doi]

Authors

Suman V. Ravuri

This author has not been identified. Look up 'Suman V. Ravuri' in Google

Steven Wegmann

This author has not been identified. Look up 'Steven Wegmann' in Google