Efficient Combination of Likelihood Recycling and Batch Calculation for Fast Acoustic Likelihood Calculation

Efficient Combination of Likelihood Recycling and Batch Calculation for Fast Acoustic Likelihood Calculation

This paper proposes an efficient combination of state likelihood recycling and batch state likelihood calculation for accelerating acoustic likelihood calculation in an HMM-based speech recognizer. Recycling and batch calculation are each based on different technical approaches, i.e. the former is a purely algorithmic technique while the latter fully exploits computer architecture. To accelerat...

Using SIMD instructions for fast likelihood calculation in LVCSR

Rapid Likelihood Calculation of Subspace

Likelihood Calculation for a Class of Multiscale Stochastic Models

A class of multiscale stochastic models based on scale-recursive dynamics on trees has recently been introduced. Theoretical and experimental results have shown that these models provide an extremely rich framework for representing both processes which are intrinsically multiscale, e.g., 1=f processes, as well as 1-D Markov processes and 2-D Markov random elds. Moreover, eecient optimal estimat...

A Bayesian Approach for Likelihood Calculation in Genetic Linkage

Monte Carlo methods play an important role in new statistical inference. Genetic linkage analysis involves estimating parameters in a genetic model in which a genetic trait is regressed on some factors such as polygenic values and environmental effects. Since only phenotypes are observed, hypothesis testing in such cases needs calculation of likelihood function in which one needs to consider al...