Improved noise-robustness in distributed speech recognition via perceptually-weighted vector quantisation of filterbank energies

There are no files associated with this record.

Title Improved noise-robustness in distributed speech recognition via perceptually-weighted vector quantisation of filterbank energies
Author So, Stephen; Paliwal, Kuldip Kumar
Publication Title Interspeech 2005 - Eurospeech
Editor Luís Oliveira
Year Published 2005
Place of publication Lisbon, Portugal
Publisher International Speech Communication Association (ISCA)
Abstract In this paper, we examine a coding scheme for quantising feature vectors in a distributed speech recognition environment that is more robust to noise. It consists of a vector quantiser that operates on the logarithmic filterbank energies (LFBEs). Through the use of a perceptually-weighted Euclidean distance measure, which emphasises the LFBEs that represent the spectral peaks, the vector quantiser codebook provides \emph{a priori} knowledge of the spectral characteristics of clean speech and is used to quantise features from noise-corrupted speech. Our comparative results from the ETSI Aurora-2 recognition task show that the perceptually-weighted vector quantisation of LFBEs achieves higher recognition accuracies for noisy speech than the unweighted vector quantisation, memoryless and multi-frame GMM-based block quantisation and scalar quantisation of Mel frequency-warped cepstral coefficients.
Peer Reviewed Yes
Published Yes
ISBN 10184074
Conference name 9th European Conference on Speech Communication and Technology
Location Lisbon, Portugal
Date From 2005-09-04
Date To 2005-09-08
URI http://hdl.handle.net/10072/2673
Date Accessioned 2006-02-22
Language en_AU
Research Centre Institute for Integrated and Intelligent Systems
Faculty Faculty of Engineering and Information Technology
Subject Signal Processing; Speech Recognition
Publication Type Conference Publications (Full Written Paper - Refereed)
Publication Type Code e1

Show simple item record

Griffith University copyright notice