DHMM speech recognition algorithm based on immune particle swarm vector quantization

Authors:
Aiping Ning;Xueying Zhang;Wei Duan
Affiliations:
College of Information Engineering, Taiyuan University of Technology, Taiyuan, China and College of Electronics and Information Engineering, Taiyuan University of Science and Technology, Taiyuan, ...;College of Information Engineering, Taiyuan University of Technology, Taiyuan, China;College of Information Engineering, Taiyuan University of Technology, Taiyuan, China
Venue:
AICI'11 Proceedings of the Third international conference on Artificial intelligence and computational intelligence - Volume Part III
Year:
2011

Citing 4
Cited 0

Swarm intelligence

Swarm intelligence
Application areas of AIS: The past, the present and the future

Applied Soft Computing
The synergy between bounded-distance HMM and spectral subtraction for robust speech recognition

Speech Communication
Image compression method using improved PSO vector quantization

ICNC'05 Proceedings of the First international conference on Advances in Natural Computation - Volume Part III

Quantified Score

Hi-index	0.00

Visualization

Abstract

This article presents a novel Immune Particle Swarm Optimization (IPSO), which combines the artificial immune system methods like immunologic memory, immunologic selection and vaccination together, by making reference to the self adjusting mechanism derived from biological immune system. IPSO as a method of Vector Quantization applied to the Discrete Hidden Markov Model (DHMM) and proposes IPSO-DHMM speech recognition algorithm. Each particle represents a codebook in the algorithm. The experiments using IPSO vector quantization algorithm get optimal codebook. Finally it enters the DHMM speech recognition system to train and recognize. The experimental results show that the IPSO-DHMM speech recognition system has faster convergence, higher recognition ratio and better robustness than the PSO-DHMM algorithm.