International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 44 - Issue 14 |
Published: April 2012 |
Authors: M. Mathivanan, S.Chenthur Pandian |
![]() |
M. Mathivanan, S.Chenthur Pandian . Efficient Speech Enhancement Approach based on Minima Controlled Recursive Averaging through Modified Map Criterion using Hidden Markov Model. International Journal of Computer Applications. 44, 14 (April 2012), 27-34. DOI=10.5120/6333-8708
@article{ 10.5120/6333-8708, author = { M. Mathivanan,S.Chenthur Pandian }, title = { Efficient Speech Enhancement Approach based on Minima Controlled Recursive Averaging through Modified Map Criterion using Hidden Markov Model }, journal = { International Journal of Computer Applications }, year = { 2012 }, volume = { 44 }, number = { 14 }, pages = { 27-34 }, doi = { 10.5120/6333-8708 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2012 %A M. Mathivanan %A S.Chenthur Pandian %T Efficient Speech Enhancement Approach based on Minima Controlled Recursive Averaging through Modified Map Criterion using Hidden Markov Model%T %J International Journal of Computer Applications %V 44 %N 14 %P 27-34 %R 10.5120/6333-8708 %I Foundation of Computer Science (FCS), NY, USA
Speech coding has become one of the most essential techniques in telecommunications and in the multimedia infrastructure. Existing speech coding techniques are applicable only for stationary environment and degrade the speech quality. This paper proposes a novel speech coding technique with better speech quality through MCRA and modified MAP. Maximum A Posteriori (MAP) criterion is extensively utilized in the statistical model-based Minima Controlled Recursive Averaging (MCRA) approaches. In the traditional MAP criterion, the inter-frame correlation of the voice activity is not taken into account. A novel technique to enhance the MCRA depending on the modified MAP via two-state Hidden Markov Model (HMM) is presented in this paper. With the proposed MAP criterion, the decision rule is obtained by clearly integrating the a priori, a posteriori, and inter-frame correlation information into the Likelihood Ratio Test (LRT).