Using Maximum Likelihood Technique to Classify and Recognized the Voices of Spoken Words

Abstract

The recognition system has two elementary phases, training phase and classification and matching phase, So we are concentrate to use classification technique to get good result, Therefore we are using maximum likelihood technique as a pattern for classification, also this model will update and enhanced in a form, we can be use it more effective in speech recognition system, the average of experimental results shown that the percentage of recognition of proposed system is about (78.57%). Also the likelihood of specific class has high values, this indicated that the system is efficient in classification process