A speech recognition system implements the task of automatically transcribing speech into text. As computer power has advanced and sophisticated tools have become available, there has been significant progress in this field. But a huge gap still exists between the performance of the Automatic Speech Recognition (ASR) systems and human listeners. In this thesis, a novel signal analysis technique...

We propose the segmented iHMM (siHMM), a hierarchical infinite hidden Markov model (iHMM) that supports a simple, efficient inference scheme. The siHMM is well suited to segmentation problems, where the goal is to identify points at which a time series transitions from one relatively stable regime to a new regime. Conventional iHMMs often struggle with such problems, since they have no mechanis...

Hidden Markov random fields appear naturally in problems such as image segmentation, where an unknown class assignment has to be estimated from the observations at each pixel. Choosing the probabilistic model that best accounts for the observations is an important first step for the quality of the subsequent estimation and analysis. A commonly used selection criterion is the Bayesian Informatio...

We present a new brain segmentation framework which we apply to T1-weighted magnetic resonance image segmentation. The innovation of the algorithm in comparison to the state-of-the-art of nonsupervised brain segmentation is twofold. First, the algorithm is entirely non-parametric and non-supervised. We can therefore enhance the classically used gray level information of the images by other feat...

In this paper, the activity recognition problem from 3-d acceleration data measured with body-worn accelerometers is formulated as a problem of multidimensional time series segmentation and classification. More specifically, the proposed approach uses a statistical model based on Multiple Hidden Markov Model Regression (MHMMR) to automatically analyze the human activity. The method takes into a...

Motivated by Hubert's segmentation procedure [16, 17], we discuss the application of hidden Markov models (HMM) to the segmentation of hydrological and enviromental time series. We use a HMM algorithm which segments time series of several hundred terms in a few seconds and is compu-tationally feasible for even longer time series. The segmentation algorithm computes the Maximum Likelihood segmen...

We propose a Tabu search based Expectation Maximization (EM) algorithm for image segmentation in an unsupervised frame work. Hidden Markov Random Field (HMRF) model is used to model the images. The observed image is considered to be a realization of Gaussian Hidden Markov Random Field (GHMRF) model. The segmentation problem is formulated as a pixel labeling problem. The GHMRF model parameters a...