Bayesian Speaker Adaptation Based on a New Hierarchical Probabilistic Model

• Published in: IEEE transactions on audio, speech, and language processing Vol. 20; no. 7; pp. 2002 - 2015
• Main Authors: Wen-Lin Zhang, Wei-Qiang Zhang, Bi-Cheng Li, Dan Qu, Johnson, M. T.
• Format: Journal Article
• Language: English
• Published: Piscataway, NJ IEEE 01.09.2012; Institute of Electrical and Electronics Engineers
• Subjects: Adaptation; Adaptation models; Applied sciences; Bayesian analysis; Correlation; Eigenphones; eigenvoices; Exact sciences and technology; Hidden Markov models; hierarchical model; Information, signal and communications theory; Mathematical models; maximum a posteriori (MAP); Principal component analysis; Probabilistic logic; Probability theory; Signal processing; speaker adaptation; Speech; Speech processing; Subspaces; Telecommunications and information theory; Telephones; Training; Vectors; Performance evaluation; Probabilistic approach; Linear regression; Subspace method; Speaker adaptation; Algorithm; Modeling; Eigenphones; Statistical method; maximum a posteriori (MAP); A posteriori estimation; Batch process; Speech recognition; On line processing; Chinese; Probabilistic model; hierarchical model; Maximum likelihood; Bayes methods; Hierarchical system; Speech processing; eigenvoices; Principal component analysis
• ISSN: 1558-7916; 1558-7924
• DOI: 10.1109/TASL.2012.2193390

In this paper, a new hierarchical Bayesian speaker adaptation method called HMAP is proposed that combines the advantages of three conventional algorithms, maximum a posteriori (MAP), maximum-likelihood linear regression (MLLR), and eigenvoice, resulting in excellent performance across a wide range of adaptation conditions. The new method efficiently utilizes intra-speaker and inter-speaker correlation information through modeling phone and speaker subspaces in a consistent hierarchical Bayesian way. The phone variations for a specific speaker are assumed to be located in a low-dimensional subspace. The phone coordinate, which is shared among different speakers, implicitly contains the intra-speaker correlation information. For a specific speaker, the phone variation, represented by speaker-dependent eigenphones, are concatenated into a supervector. The eigenphone supervector space is also a low dimensional speaker subspace, which contains inter-speaker correlation information. Using principal component analysis (PCA), a new hierarchical probabilistic model for the generation of the speech observations is obtained. Speaker adaptation based on the new hierarchical model is derived using the maximum a posteriori criterion in a top-down manner. Both batch adaptation and online adaptation schemes are proposed. With tuned parameters, the new method can handle varying amounts of adaptation data automatically and efficiently. Experimental results on a Mandarin Chinese continuous speech recognition task show good performance under all testing conditions.