Gaussian Mixture Model (GMM) computation is known to be one of the most computation-intensive components of speech recognition. In our previous work, context-independent model based GMM selection (CIGMMS) was found to be an effective way to reduce the cost of GMM computation without significant loss in recognition accuracy. In this work, we propose three methods to further improve the performan...

This paper addresses the problem of fitting finite Gaussian Mixture Model (GMM) with unknown number of components to the univariate and multivariate data. The typical method for fitting a GMM is Expectation Maximization (EM) in which many challenges are involved i.e. how to initialize the GMM, how to restrict the covariance matrix of a component from becoming singular and setting the number of ...

To reduce the high dimensionality required for training of feature vectors in speaker identification, we propose an efficient GMM based on local PCA with fuzzy clustering. The proposed method firstly partitions the data space into several disjoint clusters by fuzzy clustering, and then performs PCA using the fuzzy covariance matrix on each cluster. Finally, the GMM for speaker is obtained from ...

The chapter discusses generalized method of moments (GMM) estimation methods for spatial models. Much of the discussion is on GMM estimation of Cliff-Ord-type models where spatial interactions are modeled in terms of spatial lags. The chapter also discusses recent developments on GMM estimation from data processes which are spatially α-mixing. I.R. Prucha Department of Economics, University of ...

In the analysis of musical acoustics, we usually use the power spectrum to describe the difference between timbres from two music instruments. However, according to our experiments, the power spectrum cannot be used as effective features for erhu instrument identification. In this paper, we use MFCC (mel-scale frequency cepstral coefficients) as features for music instrument identification usin...

Methods for missing-feature reconstruction substitute noisecorrupted features with clean-speech estimates calculated based on reliable information found in the noisy speech signal. Gaussian mixture model (GMM) based reconstruction has conventionally focussed on reliable information present in a single frame. In this work, GMM-based reconstruction is applied on windows that span several time fra...

ABSTRACT: To solve the problems of outliers and high dimensionality of training feature vectors in speaker identification, in this paper, we propose an efficient GMM based on local robust PCA with VQ. The proposed method firstly partitions the data space into several disjoint regions by VQ, and then performs robust PCA using the iteratively reweighted covariance matrix in each region. Finally, ...

Estimation of the dynamic error components model is considered using two alternative linear estimators that are designed to improve the properties of the standard firstdifferenced GMM estimator. Both estimators require restrictions on the initial conditions process. Asymptotic efficiency comparisons and Monte Carlo simulations for the simple AR(1) model demonstrate the dramatic improvement in p...

Dynamic Frequency Warping (DFW) offers an appealing alternative to GMM-based voice conversion, which suffers from ”over-smoothing” that hinders speech quality. However, to adjust spectral power after DFW, previous work returns to GMMtransformation. This paper proposes a more effective DFWwith amplitude scaling (DFWA) that functions on the acoustic class level and is independent of GMM-transform...

This paper essentially extends the generalized spectral estimation of Berkowitz (2001) to provide a consistent generalized spectral estimator (GSE), considering all the information available, possibly with in…nite dimensions, based upon Escanciano (2006). Our estimator can entertain the strengths of the Berkowitz-GSE over the standard GMM. In contrast, more importantly, the newly proposed estim...