Abstract
We propose a new feature normalization scheme based on eigenspace, for achieving robust speech recognition. In particular, we employ the Mean and Variance Normalization (MVN) in eigenspace using unique and independent eigenspaces to cepstra, delta and delta-delta cepstra respectively. We also normalize training data in eigenspace and get the model from the normalized training data. In addition, a feature space rotation procedure is introduced to reduce the mismatch of training and test data distribution in noisy condition. As a result, we obtain a substantial recognition improvement over the basic eigenspace normalization.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Yao, K., Visser, E., Kwon, O., Lee, T.: A Speech Processing Front-End with Eigenspace Normalization for Robust Speech Recognition in Noisy Automobile Environments. In: Eurospeech 2003, pp. 9–12 (2003)
Molau, S., Keysers, D., Ney, H.: Matching training and test data distributions for robust speech recognition. Speech Communication 41(4), 579–601 (2003)
Jain, P., Hermansky, H.: Improved Mean and Variance Normalization for Robust Speech Recognition. In: Proc. of ICASSP (2001)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Lee, Y., Ko, H. (2004). A New Feature Normalization Scheme Based on Eigenspace for Noisy Speech Recognition. In: Apostolico, A., Melucci, M. (eds) String Processing and Information Retrieval. SPIRE 2004. Lecture Notes in Computer Science, vol 3246. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-30213-1_11
Download citation
DOI: https://doi.org/10.1007/978-3-540-30213-1_11
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-23210-0
Online ISBN: 978-3-540-30213-1
eBook Packages: Springer Book Archive