Introduction to statistical pattern recognition (2nd ed.)
Introduction to statistical pattern recognition (2nd ed.)
Digital representations of speech signals
Readings in speech recognition
The use of speech knowledge in automatic speech recognition
Readings in speech recognition
Recognition of speaker-dependent continuous speech with KEAL
Readings in speech recognition
Fundamentals of speech recognition
Fundamentals of speech recognition
Statistical methods for speech recognition
Statistical methods for speech recognition
Speech Communication - Special issue on robust speech recognition
Frequency-Warping and Speaker-Normalization
ICASSP '97 Proceedings of the 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP '97)-Volume 2 - Volume 2
The acoustic-modeling problem in automatic speech recognition
The acoustic-modeling problem in automatic speech recognition
Data-driven methods for extracting features from speech
Data-driven methods for extracting features from speech
Relevancy of time-frequency features for phonetic classification measured by mutual information
ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 01
From features extraction to strong security in mobile environment: a new hybrid system
OTM'06 Proceedings of the 2006 international conference on On the Move to Meaningful Internet Systems: AWeSOMe, CAMS, COMINF, IS, KSinBIT, MIOS-CIAO, MONET - Volume Part I
| Hi-index | 0.00 |
Feature extraction plays a major role in any form of pattern recognition. Current feature extraction methods used for automatic speech recognition (ASR) and speaker verification rely mainly on properties of speech production (modeled by all-pole filters) and perception (critical-band integration simulated by Mel/Bark filter bank). We propose to use stochastic methods for designing feature extraction methods which are trained to alleviate the unwanted variability present in speech signals. In this paper we show that such data-driven methods provide significant advantages over the conventional methods both in terms of performance of ASR and in providing understanding about the nature of speech signal. The first part of the paper investigates the suitability of the cepstral features obtained by applying discrete cosine transform on logarithmic critical-band power spectra. An alternate set of basis functions were designed by linear discriminant analysis (LDA) of logarithmic critical-band power spectra. Discriminant features extracted by these alternate basis functions are shown to outperform the cepstral features in ASR experiments. The second part of the paper discusses the relevance of non-uniform frequency resolution used by current speech analysis methods like Mel frequency analysis and perceptual linear predictive analysis. It is shown that LDA of the short-time Fourier spectrum of speech yields spectral basis functions which provide comparatively lower resolution to the high-Frequency region of spectrum. This is consistent with critical-band resolution and is shown to be caused by the spectral properties of vowel sounds.