Training log-linear acoustic models in higher-order polynomial feature space for speech recognition

Abstract

The use of higher-order polynomial acoustic features can improve the performance of automatic speech recognition. However, the dimensionality of the polynomial representation can be prohibitively large, making the training of acoustic models using polynomial features for large vocabulary ASR systems infeasible. This paper presents an iterative polynomial… (More)

Topics

Cite this paper

@inproceedings{Tahir2013TrainingLA, title={Training log-linear acoustic models in higher-order polynomial feature space for speech recognition}, author={Muhammad Ali Tahir and Heyun Huang and Ralf Schl{\"{u}ter and Hermann Ney and Louis ten Bosch and Bert Cranen and Lou Boves}, booktitle={INTERSPEECH}, year={2013} }