https://www.cs.ubc.ca/~murphyk/Teaching/CS540-Spring10/projects/Yu-hsmm09.pdf

Hidden semi-Markov models

Hidden semi-Markov models (HSMMs) are among the most important models in the area of artificial intelligence / machine learning. Since the first HSMM was introduced in 1980 for machine recognition of speech, three other HSMMs have been proposed, with various definitions of duration and observation distributions. Those models have different expressions, algorithms, computational complexities, and applicable areas, without explicitly interchangeable forms. Hidden Semi-Markov Models: Theory, Algorithms and Applications provides a unified and foundational approach to HSMMs, including various HSMMs (such as the explicit duration, variable transition, and residential time of HSMMs), inference and estimation algorithms, implementation methods and application instances. Learn new developments and state-of-the-art emerging topics as they relate to HSMMs, presented with examples drawn from medicine, engineering and computer science.Discusses the latest developments and emerging topics in the field of HSMMsIncludes a description of applications in various areas including, Human Activity Recognition, Handwriting Recognition, Network Traffic Characterization and Anomaly Detection, and Functional MRI Brain Mapping.Shows how to master the basic techniques needed for using HSMMs and how to apply them.

Hidden Semi-Markov Models: Theory, Algorithms and Applications

A major cause of degradation of speech quality in HMMbased speech synthesis is the use of a simple delta pulse signal to generate the excitation of voiced speech. This paper describes a new approach to using an acoustic glottal source model in HMM-based synthesisers. The goal is to improve speech quality and parametric flexibility to better model and transform voice characteristics.

/pdf/hmm-based-speech-synthesis-using-an-acoustic-glottal-source-4kcgf4tk1q.pdf

HMM-based Speech Synthesis Using an Acoustic Glottal Source Model

We develop a sticky hidden Markov model (HMM) with a Dirichlet distribution (DD) prior, motivated by the problem of analyzing comparative genomic hybridization (CGH) data. As formulated the sticky DD-HMM prior is employed to infer the number of states in an HMM, while also imposing state persistence. The form of the proposed hierarchical model allows efficient variational Bayesian (VB) inference, of interest for large-scale CGH problems. We compare alternative formulations of the sticky HMM, while also examining the relative efficacy of VB and Markov chain Monte Carlo (MCMC) inference. To validate the formulation, example results are presented for an illustrative synthesized data set and our main application-CGH, for which we consider data for breast cancer. For the latter, we also make comparisons and partially validate the CGH analysis through factor analysis of associated (but distinct) gene-expression data.

/pdf/sticky-hidden-markov-modeling-of-comparative-genomic-gleb47uxoo.pdf

Sticky Hidden Markov Modeling of Comparative Genomic Hybridization

In this work, the amplitude profile of sampled speech data were fitted by employing sum of sine functions with a confidence level more than 90%. Furthermore, amplitude correlation technique is applied between original speech signal samples of normal and pathological subjects and correlation technique is also applied between the curve fit constant values for normal and pathological subjects. Results obtained in both the techniques were compared to determine the varying degrees of speech disability severity.

Optimal Curve Fitting of Speech Signal for Disabled Children

Hidden Markov models (HMMs) have been used to model speech in many areas of speech processing. One characteristic of the HMM is that the probability of time spent in a particular state, or state occupancy, is geometrically distributed. This, however, becomes a serious limitation and results in inaccurate modeling when the HMMs are used for phoneme recognition. The authors use hidden semi-Markov models (HSMM) to overcome the above limitation. Semi-Markov models are a more general class of Markov chains in which the state occupancy can be explicitly modeled by an arbitrary probability mass distribution. The authors use non-parametric distributions to describe the state occupancies instead of parametric distributions such as gamma. Poisson or binomial, as analysis of actual data shows that the duration of some phonemes could not be approximated by any of the above. Preliminary tests conducted using only the linear prediction coding (LPC) cepstrum as features have shown that the use of HSMM increased the phoneme recognition accuracy to 53.7% from the 48.4% obtained using an HMM. >

N. Ratnayake

Papers

Use of semi-Markov models for speaker-independent phoneme recognition