International Journal of Emerging Technology and Advanced Engineering (IJETAE)
This paper presents a speech-model using the Linear Predictive (LP) residual signal and Maximum Likelihood Estimator (MLE). With this model an accuracy of the reverberation time estimation can be improved. During past decade, the reverberation time estimation was performed using only maximum likelihood detector, which resulted in excess time of estimation. For the purpose of estimating room acoustics, it is very essential to calculate the reverberation time more accurately. The study shows that Direct-to-Reverberation Ratio can also be calculated in accordance to the estimation of Reverberation Time. These two parameters can be made useful in the application of hearing aid.