Kernel eigenvoice speaker adaptation

Mak, Brian; Kwok, James T.; Ho, Simon

doi:10.1109/tsa.2005.851971

Cited by 41 publications

(46 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other approaches have used probabilistic PCA (PPCA) in eigenvoice adaptation [34], [37]- [39]. Mak et al applied non-linear PCA using kernel methods [53], [54]. Tanji et al explored the way to efficiently cluster the speaker-phone matrix [88].…”

Section: Eigenvoicementioning

confidence: 99%

Acoustic Model Adaptation for Speech Recognition

Sairyo

2010

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYStatistical speech recognition using continuous-density hidden Markov models (CDHMMs) has yielded many practical applications. However, in general, mismatches between the training data and input data significantly degrade recognition accuracy. Various acoustic model adaptation techniques using a few input utterances have been employed to overcome this problem. In this article, we survey these adaptation techniques, including maximum a posteriori (MAP) estimation, maximum likelihood linear regression (MLLR), and eigenvoice. We also present a schematic view called the adaptation pyramid to illustrate how these methods relate to each other. key words: speech recognition, acoustic model adaptation, hidden Markov models

show abstract

Section: Eigenvoicementioning

confidence: 99%

Acoustic Model Adaptation for Speech Recognition

Sairyo

2010

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

show abstract

“…In [52], linear transforms were used to represent each cluster for CAT, also referred to as Eigen-MLLR [24] when used in an Eigenvoice-style approach. Kernelised versions of both EigenVoices [113] and Eigen-MLLR [112] have also been studied. Finally, the use of discriminative criteria to obtain the cluster parameters has been derived [193].…”

Section: Maximum a Posteriori (Map) Adaptation 251mentioning

confidence: 99%

The Application of Hidden Markov Models in Speech Recognition

Gales

Young

2007

FNT in Signal Processing

427

155

View full text Add to dashboard Cite

Hidden Markov Models (HMMs) provide a simple and effective framework for modelling time-varying spectral vector sequences. As a consequence, almost all present day large vocabulary continuous speech recognition (LVCSR) systems are based on HMMs.Whereas the basic principles underlying HMM-based LVCSR are rather straightforward, the approximations and simplifying assumptions involved in a direct implementation of these principles would result in a system which has poor accuracy and unacceptable sensitivity to changes in operating environment. Thus, the practical application of HMMs in modern systems involves considerable sophistication.The aim of this review is first to present the core architecture of a HMM-based LVCSR system and then describe the various refinements which are needed to achieve state-of-the-art performance. These refinements include feature projection, improved covariance modelling, discriminative parameter estimation, adaptation and normalisation, noise compensation and multi-pass system combination. The review concludes with a case study of LVCSR for Broadcast News and Conversation transcription in order to illustrate the techniques described.

show abstract

“…Eigenvoices employs eigen (principal component) analysis [8] to identify a set of orthogonal basis vectors. Other extended approaches have been proposed based on eigenspace such as eigen-MLLR [9], eigenspace mapping [10], and kernel eigenvoice [11], [12]. In eigenspace-based approach, the selection of eigenvectors is not based on likelihood of the training or adaptation utterances.…”

Section: Introductionmentioning

confidence: 99%

Improved Reference Speaker Weighting Using Aspect Model

Hahm

Ohkawa

Ito

et al. 2010

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYWe propose an improved reference speaker weighting (RSW) and speaker cluster weighting (SCW) approach that uses an aspect model. The concept of the approach is that the adapted model is a linear combination of a few latent reference models obtained from a set of reference speakers. The aspect model has specific latent-space characteristics that differ from orthogonal basis vectors of eigenvoice. The aspect model is a "mixture-of-mixture" model. We first calculate a small number of latent reference models as mixtures of distributions of the reference speaker's models, and then the latent reference models are mixed to obtain the adapted distribution. The mixture weights are calculated based on the expectation maximization (EM) algorithm. We use the obtained mixture weights for interpolating mean parameters of the distributions. Both training and adaptation are performed based on likelihood maximization with respect to the training and adaptation data, respectively. We conduct a continuous speech recognition experiment using a Korean database (KAIST-TRADE). The results are compared to those of a conventional MAP, MLLR, RSW, eigenvoice and SCW. Absolute word accuracy improvement of 2.06 point was achieved using the proposed method, even though we use only 0.3 s of adaptation data. key words: speaker adaptation, aspect model, reference speaker weighting, latent reference model

show abstract

Kernel eigenvoice speaker adaptation

Cited by 41 publications

References 26 publications

Acoustic Model Adaptation for Speech Recognition

Acoustic Model Adaptation for Speech Recognition

The Application of Hidden Markov Models in Speech Recognition

Improved Reference Speaker Weighting Using Aspect Model

Contact Info

Product

Resources

About