This paper examines techniques for speaker normalisation and adaptation that are applied in training with the aim of removing some of the variability from the speaker independent models. Two techniques are examined: vocal tract normalisation (VTN) which estimates a single \vocal tract length" parameter for each speaker and then modi es the speech parameterisation accordingly and speaker adaptiv...

This paper presents a unified method for speech segmentation, speaker normalization of spectral features, and speaker adaptation of acoustic model for efficient meeting speech recognition. In the proposed method, input speech is segmented based on BIC (Bayesian Information Criterion), and compared against each speaker’s statistic in the training corpus of the acoustic model based on the BIC. Fa...

The combination of Maximum Likelihood Linear Regression (MLLR) with Maximum a posteriori (MAP) adaptation has been investigated for both the enrollment of a new speaker as well as for the asymptotic recognition rate after several hours of dictation. We show that a least mean square approach to MLLR is quite e ective in conjunction with phonetically derived regression classes. Results are presen...

This paper describes a computationally inexpensive, nonlinear feature transformation technique for rapid adaptation of a speech recognition system to new acoustic conditions. One of the advantages of the method is that it does not require any initial decoding of the adaptation data for computing the nonlinear transform. This technique performs as well as the more expensive unsupervised MLLR tec...

A method of accent modeling through Pronunciation Dictionary Adaptation (PDA) is presented. We derive the pronunciation variation between canonical speaker groups and accent groups and add an encoding of the differences to a canonical dictionary to create a new, adapted dictionary that reflects the accent characteristics. The pronunciation variation information is then integrated with acoustic ...

Speaker Adaptive Training (SAT) has been investigated for mixture density estimation and applied to large vocabulary continuous speech recognition. SAT integrates MLLR adaptation in the HMM training and aims at reducing inter-speaker variability to get enhanced speakerindependent models. Starting from BBN's work on compact models, we derive a one-pass Viterbi formulation of SAT that performs jo...

In this paper, we employ the concept of HMM-Sufficient Statistics (HMM-Suff Stat) and N-best speakers selection to realize a rapid implementation of Baum-Welch and MLLR. Only a single arbitrary utterance is required which is used to select the N-best speakers HMM-Suff Stat from the training database as adaptation data. Since HMM-Suff Stat are pre-computed offline, computation load is minimized....

The Cdc28 protein kinase subunits, Cks1 and Cks2, play dual roles in Cdk-substrate specificity and Cdk-independent protein degradation, in concert with the E3 ubiquitin ligase complexes SCFSkp2 and APCCdc20. Notable targets controlled by Cks include p27 and Cyclin A. Here, we demonstrate that Cks1 and Cks2 proteins interact with both the MllN and MllC subunits of Mll1 (Mixed-lineage leukaemia 1...

The performance of global affine and nonlinear transformations for speaker adaptation in a hidden Markov model (HMM) speech recognition system are compared in this paper. The nonlinear transformation was obtained with a multilayer perceptron network (MLP) which was trained during the adaptation process to transform the mean vectors of the HMMs such that the output probabilities of the HMMs for ...

Eigen-MLLR coe cients are proposed as new feature parameters for speaker-identi cation in this paper. By performing principle component analysis on MLLR parameters among training speakers, the eigen-MLLR coe cients (EMCs) are derived as the coe cients for the eigenvectors. The discriminating function of the new EMC features based on the Fisher criterion is found to be ten times larger than that...