Paper Info
Title
Maximum a posteriori adaptation of HMM parameters based on speaker space projection
Abstract
This paper presents a novel approach to rapid speaker adaptation based on the speaker space projection paradigm in which the adapted model is constrained to lie on a specific subspace spanned by a small number of basis vectors. In order to select the basis vectors that form the speaker space, we apply probabilistic principal component analysis (PPCA) technique to a set of training speaker models represented by a number of hidden Markov models (HMMs). The PPCA incorporates a probability model to the conventional principal component analysis (PCA) method, and finds the speaker space model by means of the expectation maximization (EM) algorithm which is computationally efficient. The PPCA model provides the information of correlation among different speech units as well as the prior probability density function (pdf) associated with each HMM parameter, which can be directly applied to the maximum a posteriori (MAP) adaptation framework. Through a series of supervised adaptation experiments on the tasks of connected digit and large vocabulary recognition, we show that the proposed approach not only achieves a good performance for a small amount of adaptation data but also guarantees a consistent estimate as the data size grows.
Year
DOI
Venue
2004
10.1016/j.specom.2003.08.001
Speech Communication
Keywords
Field
DocType
Hidden Markov model,Maximum a posteriori,Speaker adaptation,Probabilistic principal component analysis,EM algorithm
Pattern recognition,Subspace topology,Computer science,Expectation–maximization algorithm,Speech recognition,Artificial intelligence,Maximum a posteriori estimation,Estimation theory,Prior probability,Hidden Markov model,Basis (linear algebra),Principal component analysis
Journal
Volume
Issue
ISSN
42
1
0167-6393
Citations 
PageRank 
References 
3
0.40
14
Authors
2
Name
Order
Citations
PageRank
Dong Kook Kim1509.44
Nam Soo Kim227529.16