Using Syllable Nuclei Locations To Improve Automatic Speech Recognition In The Presence Of Burst Noise

In this work we combine a conventional phone-based automatic speech recognizer with a classifier that detects syllable locations. This is done using a dynamic Bayesian network. Using oracle syllable detections we achieve a 17% relative reduction in word error rate on the 500 word task of the SVitchboard corpus. Using estimated locations we achieve a 2.1% relative reduction which is significant at the 0.02 level. The improvement in the estimated case is from reducing insertions caused by burst noise.