This paper presents an eecient approximation of the Gaussian mixture state probability density functions of continuous observation density hidden Markov models (CHMM's). In CHMM's, the Gaussian mixtures carry a high computational cost, which amounts to a signii-cant fraction (e.g. 30% to 70%) of the total computation. To achieve higher computation and memory ee-ciency, we approximate the Gaussi...

This report explains the theory of Hidden Markov Models (HMMs). The emphasis is on the theory aspects in conjunction with the implementation issues that are encountered in a floating point processor. The main theory and implementation issues are based on the use of a Gaussian Mixture Model (GMM) as the state density in the HMM, and a Continuous Density Hidden Markov Model (CDHMM) is assumed. Su...

In this study Hidden Markov models are employed for the automatic classiication of pitch accents in German utterances. A sim-pliied version of the tone sequence model (i.e. a linguistic theory of pitch accents) is applied. In this approach only two of the nuclear tones (rise, fall) are used. They are represented by continuous density Hidden Markov models. The classiier works on the sequence of ...

Hidden Markov models and their variants are the predominant sequential classification method in such domains as speech recognition, bioinformatics and natural language processing. Being generative rather than discriminative models, however, their classification performance is a drawback. In this paper we apply ideas from the field of density ratio estimation to bypass the difficult step of lear...

In this paper, we introduce several hybrid connectionist-structural acoustic models for contextindependent phone-like units in the atros recognition system. The structural part of the acoustic models has been modeled with Markov chains, and a multilayer perceptron (or a committee of multilayer perceptrons) is used to estimate the emission probabilities of the Markov chains. We compare the recog...

Speech Recognition is a process of transcribing speech to text. Phoneme based modeling is used where in each phoneme is represented by Continuous Density Hidden Markov Model. Mel Frequency Cepstral Coefficients (MFCC) are extracted from speech signal, delta and double-delta features representing the temporal rate of change of features are added which considerably improves the recognition accura...

A semi-continuous hidden Markov model based on the multiple vector quantization codebooks is used here for large-vocabulary speaker-independent continuous speech recognition. In the techniques employed here, the semi-continuous output probability density function for each codebook is represented by a combination of the corresponding discrete output probabilities of the hidden Markov model and t...

We study a two-dimensional discounted optimal stopping problem related to the pricing of perpetual commodity equities in model financial markets which behaviour underlying asset price follows generalized geometric Brownian motion and dynamics convenience yield are described by an unobservable continuous-time Markov chain with two states. It is shown that time exercise first at spot paid return ...