A synergy between HMM-GA based on stochastic cellular automata to accelerate speech recognition

Most current speech recognition systems are based on Hidden Markov Model (HMM) using Viterbi search in decoding stage. Whereas such algorithm is in terms of dynamic programming, it consists of many computations with increasing number of reference words. This paper presents a novel method called HGSCA computing likelihood measurement between an unknown input pattern and a reference model based on a synergy between HMM and Genetic Algorithm (GA) in a parallel form by Stochastic Cellular Automata (SCA). The HGSCA algorithm is compared with the Viterbi algorithm from the “recognition time” and “recognition error” view points. Experimental results show although HGSCA and Viterbi algorithms are very close in recognition accuracy, the HGSCA is so faster than the Viterbi algorithm.

[1]  A. E. Eiben,et al.  Introduction to Evolutionary Computing , 2003, Natural Computing Series.

[2]  Biing-Hwang Juang,et al.  Fundamentals of speech recognition , 1993, Prentice Hall signal processing series.

[3]  C. D. Gelatt,et al.  Optimization by Simulated Annealing , 1983, Science.

[4]  Oliver Chiu-sing Choy,et al.  A Speech Recognition IC Using Hidden Markov Models with Continuous Observation Densities , 2007, J. VLSI Signal Process..

[5]  Yaneer Bar-Yam,et al.  Dynamics Of Complex Systems , 2019 .

[6]  B. Venkataramani,et al.  SOC Implementation of HMM Based Speaker Independent Isolated Digit Recognition System , 2007, 20th International Conference on VLSI Design held jointly with 6th International Conference on Embedded Systems (VLSID'07).

[7]  Marco Gori,et al.  A survey of hybrid ANN/HMM models for automatic speech recognition , 2001, Neurocomputing.

[8]  Saeed Setayeshi,et al.  Accelerating Speech Recognition Algorithm with Synergic Hidden Markov Model and Genetic Algorithm Based on Cellular Automata , 2009, 2009 International Conference on Signal Processing Systems.