Comparison of Speaker Adaptation Methods as Feature Extraction for SVM-Based Speaker Recognition

• Published in: IEEE transactions on audio, speech, and language processing Vol. 18; no. 6; pp. 1366 - 1378
• Main Authors: Ferras, M, Cheung-Chi Leung, Barras, C, Gauvain, J.-L.
• Format: Journal Article
• Language: English
• Published: Piscataway, NJ IEEE 01.08.2010; Institute of Electrical and Electronics Engineers
• Subjects: Adaptation model; Applied sciences; Computer Science; Constrained MLLR (CMLLR); Exact sciences and technology; Feature extraction; Gaussian mixture model (GMM); Gaussian supervectors; Information theory; Information, signal and communications theory; Loudspeakers; Maximum likelihood estimation; Maximum likelihood linear regression; maximum-likelihood linear regression (MLLR); Miscellaneous; Parameter estimation; Signal and communications theory; Signal and Image Processing; Signal processing; Signal representation. Spectral analysis; Signal, noise; Speaker recognition; Speech; Speech processing; support vector machine (SVM); Support vector machine classification; Support vector machines; Telecommunications and information theory; Performance evaluation; Generalized steering vector; Automatic classification; support vector machine (SVM); Mixture theory; Speaker adaptation; Support vector machine; Learning; User interface; A posteriori estimation; maximum-likelihood linear regression (MLLR); Audio signal processing; Linear regression; Man machine dialogue; Speaker recognition; Acoustic signal processing; Signal classification; Multiple models approach; Gaussian process; Gaussian super-vectors; Information processing; Signal processing; Constrained MLLR (CMLLR); Feature extraction; Data fusion; Gaussian mixture model (GMM); Maximum likelihood; Speech processing; MLLR; SVM; CMLLR; GMM; Gaussian supervectors
• ISSN: 1558-7916; 1558-7924
• DOI: 10.1109/TASL.2009.2034187

In the last years the speaker recognition field has made extensive use of speaker adaptation techniques. Adaptation allows speaker model parameters to be estimated using less speech data than needed for maximum-likelihood (ML) training. The maximum a posteriori (MAP) and maximum-likelihood linear regression (MLLR) techniques have typically been used for adaptation. Recently, MAP and MLLR adaptation have been incorporated in the feature extraction stage of support vector machine (SVM)-based speaker recognition systems. Two approaches to feature extraction use a SVM to classify either the MAP-adapted Gaussian mean vector parameters (GSV-SVM) or the MLLR transform coefficients (MLLR-SVM). In this paper, we provide an experimental analysis of the GSV-SVM and MLLR-SVM approaches. We largely focus on the latter by exploring constrained and unconstrained transforms and different choices of the acoustic model. A channel-compensated front-end is used to prevent the MLLR transforms to adapt to channel components in the speech data. Additional acoustic models were trained using speaker adaptive training (SAT) to better estimate the speaker MLLR transforms. We provide results on the NIST 2005 and 2006 Speaker Recognition Evaluation (SRE) data and fusion results on the SRE 2006 data. The results show that using the compensated front-end, SAT models and multiple regression classes bring major performance improvements.