Low-Variance Multitaper MFCC Features: A Case Study in Robust Speaker Verification
Author
Summary, in English
In speech and audio applications, short-term signal spectrum is often represented using mel-frequency cepstral coefficients (MFCCs) computed from a windowed discrete Fourier transform (DFT). Windowing reduces spectral leakage but variance of the spectrum estimate remains high. An elegant extension to windowed DFT is the so-called multitaper method which uses multiple time-domain windows (tapers) with frequency-domain averaging. Multitapers have received little attention in speech processing even though they produce low-variance features. In this paper, we propose the multitaper method for MFCC extraction with a practical focus. We provide, first, detailed statistical analysis of MFCC bias and variance using autoregressive process simulations on the TIMIT corpus. For speaker verification experiments on the NIST 2002 and 2008 SRE corpora, we consider three Gaussian mixture model based classifiers with universal background model (GMM-UBM), support vector machine (GMM-SVM) and joint factor analysis (GMM-JFA). Multitapers improve MinDCF over the baseline windowed DFT by relative 20.4% (GMM-SVM) and 13.7% (GMM-JFA) on the interview-interview condition in NIST 2008. The GMM-JFA system further reduces MinDCF by 18.7% on the telephone data. With these improvements and generally noncritical parameter selection, multitaper MFCCs are a viable candidate for replacing the conventional MFCCs.
Department/s
- Mathematical Statistics
- Statistical Signal Processing Group
Publishing year
2012
Language
English
Pages
1990-2001
Publication/Series
IEEE Transactions on Audio, Speech, and Language Processing
Volume
20
Issue
7
Document type
Journal article
Publisher
IEEE - Institute of Electrical and Electronics Engineers Inc.
Topic
- Probability Theory and Statistics
Keywords
- Mel-frequency cepstral coefficient (MFCC)
- multitaper
- small-variance
- estimation
- speaker verification
Status
Published
Research group
- Statistical Signal Processing Group
ISBN/ISSN/Other
- ISSN: 1558-7924