Measures the relative contributions of the key differences in the regression model, state vs. frame predictions, and separate vs. combined stream predictions.
Oliver Watts, Gustav Eje Henter, Thomas Merritt, Zhizheng Wu, Simon King. “From HMMs to DNNs: where do the improvements come from?” IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), April 2016, Shanghai, China. DOI: 10.1109/ICASSP.2016.7472730
Publisher’s version (preferred)
Publisher’s version (preferred for Edinburgh University students)
Author’s version (use only as a last resort)
