A stochastic version of Expectation Maximization algorithm for better estimation of Hidden Markov Model

Authors:
Shamsul Huda;John Yearwood;Roberto Togneri
Affiliations:
Center for Informatics and Applied Optimization, School of Information Technology and Mathematical Science (ITMS), University of Ballarat, VIC 3353, Victoria, Australia;Center for Informatics and Applied Optimization, School of Information Technology and Mathematical Science (ITMS), University of Ballarat, VIC 3353, Victoria, Australia;Center for Intelligent Information Processing, School of Electrical, Electronic and Computer Engineering, University of Western Australia, WA, Australia
Venue:
Pattern Recognition Letters
Year:
2009

Citing 11
Cited 2

Convergence of an annealing algorithm

Mathematical Programming: Series A and B
Fundamentals of speech recognition

Fundamentals of speech recognition
Learning mixture models using a genetic version of the EM algorithm

Pattern Recognition Letters
A segmental-feature HMM for continuous speech recognition based on a parametric trajectory model

Speech Communication
Genetic-Based EM Algorithm for Learning Gaussian Mixture Models

IEEE Transactions on Pattern Analysis and Machine Intelligence
Decomposing parameters of mixture Gaussian model using genetic and maximum likelihood algorithms on dental images

Pattern Recognition Letters
Evolutionary algorithms for constrained parameter optimization problems

Evolutionary Computation
A constraint-based evolutionary learning approach to the expectation maximization for optimal estimation of the hidden Markov model for speech signal modeling

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics - Special issue on human computing
Clustering in image space for place recognition and visualannotations for human-robot interaction

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Genetic Algorithms and Very Fast Simulated Reannealing: A comparison

Mathematical and Computer Modelling: An International Journal
Simulated annealing for maximum a posteriori parameter estimation of hidden Markov models

IEEE Transactions on Information Theory

Aging speech recognition with speaker adaptation techniques: Study on medium vocabulary continuous Bengali speech

Pattern Recognition Letters
Hierarchical speech-act classification for discourse analysis

Pattern Recognition Letters

Quantified Score

Hi-index	0.10

Visualization

Abstract

This paper attempts to overcome the local convergence problem of the Expectation Maximization (EM) based training of the Hidden Markov Model (HMM) in speech recognition. We propose a hybrid algorithm, Simulated Annealing Stochastic version of EM (SASEM), combining Simulated Annealing with EM that reformulates the HMM estimation process using a stochastic step between the EM steps and the SA. The stochastic processes of SASEM inside EM can prevent EM from converging to a local maximum and find improved estimation for HMM using the global convergence properties of SA. Experiments on the TIMIT speech corpus show that SASEM obtains higher recognition accuracies than the EM.